Cooperative Services in next-generation wireless networks

477

Chapter 19

Cooperative Services in next-generation wireless networks: internet of things Paradigm

Andreas P. Fatouros, Ioannis C. Fousekis, Dimitris E. Charilas, and Athanasios D. Panagopoulos

Contents19.1 Introduction ................................................................................................................... 47819.2 Internet of Things........................................................................................................... 478

19.2.1 The Concept of IoT ............................................................................................. 47819.2.2 Related Technologies .......................................................................................... 47919.2.3 Applications of IoT in Everyday Life ................................................................... 48019.2.4 Limitations and Problems of IoT ........................................................................ 48119.2.5 Benefits of Employment of IoT Technology ........................................................ 482

19.3 Cooperative Services in Next-Generation Wireless Networks ........................................ 48219.3.1 Cooperative Networks: General Concept............................................................ 48219.3.2 Cooperative Service Scenarios ............................................................................. 48519.3.3 Limitations and Problems of Cooperative Services .............................................. 48819.3.4 Motivations for Cooperation............................................................................... 489

19.4 Simulations and Discussion of Cooperative Network Scenarios ..................................... 49019.4.1 Simulation Scenarios .......................................................................................... 49019.4.2 Network Model .................................................................................................. 49019.4.3 Energy Consumption Estimation ....................................................................... 492

19.4.3.1 Cellular Link (UMTS) ......................................................................... 49219.4.3.2 Short-Range Link (Bluetooth) .............................................................. 492

19.4.4 Simulation Results .............................................................................................. 492

478 ◾ Building Next-Generation Converged Networks

19.1 introductionMobile networks systems are reaching a peak from the offered capabilities of a single technology. With the use of the same selfish networks, we may not be able to provide further improvement to meet the needs of the users. However, the general idea of cooperation is now on the table. The idea of creating a huge network where all different access technologies can be a part of it is now pos-sible thanks to cooperation without interference. Nevertheless, the development of such a network meets lots of problems along with limitations from the existing technologies. Moreover, end users may refuse to embrace these technologies because of seemingly increased costs. The main focus of this chapter is to highlight the cooperation services between different network technologies. More specifically, the chapter consists of three sections.

In the first section, a generic presentation about the Internet of things (IoT) is made along with the positive results that could succeed in people’s everyday lives. The general idea of IoT is introduced along with the already existing related technologies. Furthermore, the difficulties and limitations of IoT are discussed. Finally, an example of a model architecture of a smart home envi-ronment network with smart objects is introduced as well as several examples of IoT applications.

The second section of the chapter is devoted to cooperative services in next-generation wireless networks. With the cooperation of different access technologies, significantly enhanced services can be provided to end users. In addition, the already existing services can be supported even faster, thus creating an environment where higher QoS with lower cost can be offered. Furthermore, cooperative services scenarios are discussed along with ideas about implementation in the real world and limitations and problems of cooperative networks. Lastly, motives for users and service providers are proposed based on reputation mechanisms.

The chapter emphasizes on how cooperation can be implemented in the IoT paradigm. Because the development of IoT demands a ubiquitous network where everything can connect and com-municate at any time, the cooperation of the existing networks is the only cheap way to make it possible. In addition, to support the idea of cooperation within the frame of this study, a simula-tion project is developed. In the third section of the chapter, we describe the scenarios that have been simulated and discuss results. The latter demonstrate how significantly a cooperative services scenario can improve the quality of service (QoS) offered to end users.

19.2 internet of things19.2.1 The Concept of IoTNowadays, the majority of worldwide connections are made between humans, especially between humans connected through computers or mobile handsets. Consequently, the main communication is from human to human. In the concept of the IoT, every machine or thing

19.4.4.1 Comparison between Remote Node Assistance and Baseline Scenarios ... 49319.4.4.2 Comparison between Cooperative Broadcast and Baseline Scenarios ... 494

19.4.5 Discussion .......................................................................................................... 49519.5 Future Research Directions ............................................................................................ 49619.6 Conclusions .................................................................................................................... 496References ............................................................................................................................... 497

Cooperative Services in Next-Generation Wireless Networks ◾ 479

must be able to connect online to the Internet and communicate with other machines or things without the presence but on behalf of human beings. In addition, there is not a standard defi-nition of the IoT, but the general idea behind IoT is the necessity of giving to every entity the ability to communicate. Perhaps the best idea to describe the IoT is by adding a new dimension to the world of information and communication technologies and transforming the present con-nectivity at any time, from any place, and for anyone to connectivity for anything as illustrated in Figure 19.1 [1].

The next step is to specify the way these smart things can communicate no matter if the human is present or not. Perhaps this is the essential part of the IoT project in which every “stu-pid” machine has to become intelligent to be able to communicate. In the IoT, these things or machines are named as “smart” things. With the theoretical approach that basic function can be modeled by expanding the existing form into a machine-to-machine (M2M) communication or a thing-to-thing (T2T) communication [2,3]. Imagine that it is like an evolved form of the peer-to-peer (P2P) logic [4,5]. On the contrary, implementing that extension and creating a stable and trustworthy network is very difficult because of various limitations as will be discussed later.

19.2.2 Related TechnologiesAs was previously mentioned, all things that are able to connect online and communicate are named as smart things. To make an everyday thing smart, a whole set of algorithms and micro-processors have to be used in order to achieve the desirable result. Furthermore, one detail should be mentioned: the size of these things varies from huge things to tiny things. These tiny things cannot be fully equipped with all these microprocessors and sensors because of their size. In that case, radio frequency identification (RFID) tags can be used, which are small microchips designed for wireless data transmission and are used to help identify objects. On the other hand, in the IoT scenario, things that are equipped with well-known commercial wireless and mobile cards are able to participate.

Generally, RFID tags are attached to an antenna in a package that resembles an ordinary adhe-sive sticker and transmits data over the air in response to interrogation by an RFID reader. Their cost and size make them suitable for IoT’s smart things project. In addition, unique identification and automation are two characteristics that make RFID tags appropriate for object tagging. An RFID tag emits a unique serial number that distinguishes it among many millions of identically manufactured objects. These unique identifiers in the RFID tags can act as pointers to database

Anytime

Connect

Any place

Anything

figure 19.1 three-dimensional connectivity model.


entries containing rich transaction histories for individual items. Moreover, RFID tags are read-able without line-of-sight contact and without precise positioning. RFID readers can scan tags at rates of hundreds per second.

There are three types of RFID tags. In general, the small and inexpensive RFID tags are pas-sive. Especially passive tags do not have any on-board power source. Instead, they derive their transmission power from the signal of an interrogating reader. Some RFID tags contain batteries. There are two such types: semipassive tags, whose batteries power their circuitry when they are interrogated, and active tags, whose batteries power their transmissions. Note that active tags can initiate communication and have read ranges of more than 100 m [6].

19.2.3 Applications of IoT in Everyday LifeThe best way to understand the IoT project is by giving a general idea of IoT’s application devel-opment in people’s everyday lives. Because it is difficult to provide real evidence and proof of the definite success of the IoT project, a few examples are provided to support the idea. With the use of RFID tags and RFID readers, a whole set of functions and services can be used to help people. In addition, cooperating entities can support better capabilities and better communication between smart objects. To emphasize, cooperation is the only way to create a safer and more efficient net-work where RFID-tagged smart objects can communicate with each other, providing accurate results. For example, RFID tags on drugs and food can help people to identify the origin and the quality of them. In addition, with an Internet connection and a database, RFID tags and readers can support a full tracking system [7].

Furthermore, cars, trains, and buses—along with the roads and the rails—equipped with sen-sors, actuators, and RFID tags may provide important information to the driver and passengers about navigation and safety. Moreover, collision avoidance systems and monitoring of transpor-tation may protect drivers from hazardous materials transportation and other dangers [8]. Of course, the development of such a system is difficult because of different environmental factors, even though research is done to create a common architecture model in which every entity can communicate with every other one [9].

The best example showing the importance of the IoT project is a smart home environment project built by the Korea Institute of Industrial Technology (KITECH) to demonstrate the prac-ticability of a robot-assisted future home environment. This environment consists of smart objects with RFID tags and smart appliances with sensor network functionality. The home server that connects these smart devices maintains information for reliable services, and the service robots perform tasks in collaboration with the environment. Furthermore, an interesting architecture of such a network is proposed, one that shows the way a computer network would seem after the wide use of the RFID tags in smart object appliances as shown in Figure 19.2.

Examining the same smart home environment, we describe a scenario that intends to identify and fetch a requested object that is included in the environment in order to help the owner of the house. The target object is smart, which means it is RFID tagged. Using smart devices, such as a smart table, smart shelf, and smart bookcase, the detection of the presence of the requested item is possible. Once the RFID code of the smart item is detected, it is transferred to the home server through wireless networks. If the user of the smart network wants, for example, the position of a cup, he or she sends a command to the home server. The home server then searches for the status information of the device and sends the position data to the robot. After downloading the data from the home server, the robot moves to the place where the target object is and brings it to the user [10].


19.2.4 Limitations and Problems of IoTOf course, the IoT project does not have only benefits and positive outcomes in people’s everyday lives. Particularly, there are some problems and limitations in the development of the IoT and with the results of the installation as well. First of all, one of the major problems that must be solved is the communication range of the wireless networks. Indoors, this can be solved with the use of a router or a set of routers with wireless connectivity. But outdoors, this is much more difficult because of the time varying fading phenomena. At this point, the only solution that could be applied is the cooperation of the existing networks, which will create a ubiquitous cluster of wire-less networks [8,11]. At the same time, another limitation of the IoT’s development is an increased demand for power (energy). As a result of the use of wireless connections, two side effects occur mostly as a result of multipath fading: first, the power consumption of smart things increases because of retransmissions, and second, the lifetime of the batteries of smart objects is reduced. Consequently, the solution comes again from the possible cooperation of things in these networks by reducing the retransmitted data and with the smart management of them.

The second part of the IoT’s problems involves security, privacy, and identification concerns. The deployment of automatic communication networks of smart objects could possibly represent a danger for society. Embedded RFID tags in every object may result in intrusion into people’s personal lives. RFID tags can unknowingly be triggered to reply with their ID and other infor-mation without the permission of the users; therefore, security measures must be taken in order to prevent eavesdropping. One of the measures that can be taken is relabeling. According to this mode, users can physically alter tags to limit their data emission and obtain physical confirmation of their changed state. Nevertheless, the use of unique identifiers does not eliminate the threat of clandestine inventorying or the threat of tracking. In addition, even if relabeling can be applied in smart objects with active RFID chips, it cannot be applied in passive RFID tags. This is why the use of a minimalistic cryptography model is better. According to this model, every tag contains a small

Real environment

Sensors/actuators

Smart devicesSmart

bookcase

RFID-taggedcups

RFID-taggedbooks

RFID-taggedobjects

RFIDreaders

Smartswitches

Smarttables

Smartlamps

Smartsecuritycameras

Smartsensors

Routers Wi-Fispots

Smarthomeserver

Smartappliances

Luminancesensors

Networks

figure 19.2 architecture of a smart home environment network with smart objects.


collection of pseudonyms, which rotates by releasing a different one on each reader query. An autho-rized reader can store the full pseudonym set for a tag in advance and identify the tag consistently. On the other hand, an unauthorized reader is unable to correlate the different appearances of the same tag. The minimalist scheme can offer some resistance to corporate espionage, such as clandes-tine scanning of product stocks in retail environments [6]. Finally, cooperative security protocols that have been proposed in wireless sensor networks may find application in IoT configurations.

19.2.5 Benefits of Employment of IoT TechnologyThe benefits of IoT can be visible in every aspect of everyday life. An especially huge number of applications are possible for development, but only a small part of them is currently available. The domains and environments in which new applications could improve the quality of people’s lives are many: from home to work, from gym to traveling, at the hospital, anywhere it is possible to use the capabilities of smart objects. These environments are now equipped with objects with only primitive intelligence, often times without any communication capabilities. Giving these objects the possibility to communicate with each other and to elaborate on the information perceived from their surroundings makes them more effective by making people’s lives convenient [8].

Another part of the IoT project concerns the advantages that are created from the cooperation of heterogeneous networks. The whole point of the cooperative networks is to create a ubiquitous net-work, so everything can be able to communicate at any time. This whole process has the effect of the expansion of the existing network coverage and the increased capabilities of connectivity support. Because of this characteristic, smart objects can easily access the network. As a result, less energy is consumed for the communication of the smart objects, and thus, the lifetime maximization of their batteries is possible. Furthermore, the difficult process to support an efficient security model can be possible with the proportion of the security steps in different entities of the smart network.

19.3 Cooperative Services in next-generation wireless networks

19.3.1 Cooperative Networks: General ConceptThe meaning of the word “cooperative” derives from the very beginning of our society itself and can also be found in terms of biology. For example, people help each other in order to accomplish their tasks faster and in a more productive way. Cooperation’s outcome in computer or mobile net-works is getting better results in a faster way, wasting much less energy, time, thought, and money. The latest research approaches make great efforts to introduce the concept of cooperation into telecommunications in order to create the next-generation 4G networks. It is envisioned that these evolved networks will offer many advantages to users of mobile communications by augment-ing and optimizing the perceived QoS, the number of provided services to end users, and being friendlier to the environment. The demand for higher throughput capabilities, higher reliability, and high-speed mobility, as it is delineated by International Telecommunication Union (ITU) standards for 4G networks, requires great complexity, energy consumption, and expenditure from mobile devices and telecommunication service providers. Some of these standards are data rates up to 1 Gbps, compatibility with former generation networks, and better quality of multimedia services (e.g., video on demand, mobile TV, and high-definition TV). In addition, 4G networks use packet switching and the IP instead of the previous ones.


The envisioning of 4G networks is the idea of cooperative networks. The basic technical char-acteristics and objectives of cooperative 4G networks are tabulated in Table 19.1. The state-of-the-art for cooperative networks is the coexistence of many heterogeneous networks with different data rates, mobility capabilities, and different technical characteristics [12]. In general, there are two types of wireless mobile networks: wide and local access wireless networks. Wide access wire-less networks (WAN), such as general packet radio service (GPRS), global system for mobile com-munications (GSM), and universal mobile telecommunications system (UMTS) (or 3G), offer lower data rates than the local ones, but they support higher coverage and mobility to users. On the contrary, local access wireless networks (LANs), such as Bluetooth and WLANs, support much higher data rates with much less energy consumption comparatively, but they suffer in mobility and coverage. Thus, the idea of combining these two types of wireless networks in order to achieve the maximum gain regarding data rates, energy consumption, mobility, and coverage has emerged. The tradeoff between coverage and data rates can be minimized in that way. The different networks that are used all over the world will be converged into a wider network [12].

However, many difficulties have to be faced looking toward the implementation of 4G networks. The always-on Internet, the data transfer, the spectrum management, the very high data rates that have been prescribed by ITU, the complexity that will be introduced into the whole system, and the deficiency of IP addresses are some of these problems that scientists and network engineers have to confront and overcome. Some of these characteristics will increase energy consumption, which is one of the main factors that ITU has defined as of great importance for the next-generation networks. This fact adds additional difficulty in the design of network and mobile devices as well.

table 19.1 Characteristics of 4g Cooperative networks

Data transfer capability

• 100 Mbps (wide coverage)

• 1 Gbps (local area)

Networking • All-IP network (access and core networks)

• Plug and access network architecture

• An equal-opportunity network of networks

Connectivity • Ubiquitous, mobile, continuous

Network capacity • Tenfold that of 3G

Latency • Connection delay < 500 ms

• Transmission delay < 50 ms

Cost • Cost per bit: 1/10 to 1/100 lower than that of 3G

• Infrastructure cost: 1/10 lower than that of 3G

Connected entities • Anything to anything

4G networks key objectives

• Heterogeneity and convergence of networks, terminals, and services

• Harmonious wireless ecosystem

• Perceptible simplicity, hidden complexity

• Cooperation as one of its underlying principles


In cooperative networks, the concept of “term node” is introduced. The P2P communication between mobile devices and a base station (BS) is no longer applicable because the node (user or mobile device) can act as a relay. In such a scenario, the destination (D) receives data from both the BS/source (S) via a direct link and the relay (R) via a two-hop transmission as shown in Figure 19.3. This also means that the relay node has the ability not only to receive data from the BS for personal use but also to forward them to other nodes.

Generally speaking, the BS can send data to users through a direct link. However, this link may suffer from noise, which makes the communication difficult or impossible. Fading, long dis-tance between user and BS, high signal-to-noise ratio (SNR), and high packet loss rate are some of the problems that burden the communication. The introduction of the user node may help to overcome these problems as the link between BS and node or between node and destination user. That would make the data transfer more feasible to the destination and could improve the quality of provided services. In such a scenario, the BS can send data simultaneously both to the node and destination user, thus achieving a faster and more reliable transfer of data. This simple model can be applied to more than one cooperative terminal or network. The cooperation, however, between terminals of many heterogeneous networks may demand the development of new network proto-cols and the introduction of some more bits in the header of data messages that terminals exchange, which may reduce the achieved throughput. Nevertheless, the overall gain through cooperation in data rates, energy consumption, and coverage makes this small decrease in throughput negligible. The benefits of cooperative networks can be summarized in the following aspects:

◾ Data rates: Cooperation between neighboring nodes will significantly increase the data rate because devices will receive data from both the cellular and short-range links. In some situ-ations, the augmentation of the data rate has been double or even more [13,14].

◾ Transmission time: As the overall data rate increases, it is corollary that the needed time to transmit all the data requested by a user would be significantly decreased. Transmission time can be decreased by more than 50% [15].

◾ Capacity: A direct consequence of the last advantage is the increase in the system capacity. In cooperative networks, users tend to download only a part of the overall data information through cellular networks. As a result, they occupy this link much less, they waste far fewer resources from the system, and the overall capacity of the system can be increased.

◾ Coverage: Through the combination and cooperation of networks and mobile devices, the coverage of the system can be increased. Additionally, short-range links can transmit the information further [13,14].

Direct link

Two-hop channel

S D

R

figure 19.3 Basic cooperative transmission configuration.


◾ Energy consumption: As it is widely known, energy consumption in short-range links is extremely low compared to cellular links. In cooperative networks, users receive a great amount of information through short-range links, decreasing the overall amount of energy spent by users and BSs. However, it must be mentioned that cooperating nodes will spend energy for receiving data from BSs and sending them to other users. This is one of the major problems in the design of next-generation networks and can only be resolved by giving these nodes some motivation in order to cooperate with each other [13,14,16].

◾ QoS: All the above, in combination with the improvement of the SNR and the signal-to-interference ratio (SIR) that short-range links can offer, contribute to the decrease in the bit error ratio and probability (BER and BEP) and, consequently, in a spectacular improvement of QoS. This fact can lead to import new very demanding services for the users.

◾ Cost of cooperative services: Taking as reference either the time terminals spent to receive the data or the amount of data terminals received through cellular links, cooperative networks overmatch significantly compared to the former generations of networks. As a result, the cost is estimated to be much lower [15].

There are two possible ways to transfer data between nodes that are differentiated in the way the cooperating users download the data from the server. In the static server-aware approach, it is assumed that users can send or receive data simultaneously through the short-range link or cellular link. In this situation, the server must be aware of the amount of data that has to be sent to every user; consequently, mobile devices must send some information to the server before the procedure begins. Let us assume the existence of two cooperating users and two files in the server side, which are the parts of the whole file. The destination node informs the cooperating node to download the second part of the file while it downloads the first one, and then they exchange data through the short-range link. While installing the short-range link, every cooperating node becomes aware of what it has to download from the server. This scenario can be easily expanded to more than two terminal nodes.

The second scenario relies on a dynamic server-unaware approach that takes advantage of the HTTP/1.1 possibilities range request. In that situation, every terminal has the ability to decide which bytes of a file will download from the server through the HTTP GET requests. During the installation of the short-range link, the cooperating terminals decide which bytes each terminal will download, and then an HTTP GET request is sent to the server with the preagreed range request into the GET request. One of the advantages of this scenario is the flexibility that the server offers to users. Users can decide to download data through the best cellular links with the maximum data rate and the minimum noise and fading. Even if the short-range link becomes unavailable for some reason, the terminal will send another HTTP GET request to the server in order to receive the rest of the file [15].

19.3.2 Cooperative Service ScenariosObserving the 4G networks from a user’s point of view, a multitude of new applications and ser-vices is supposed to emerge. The new communication system, although cellular, is not completely based on the BS–terminal communication, but is highly dependent on short-range communica-tions. The cooperation between nodes improves the reliability of the offered services, proliferating simultaneously the coverage and data rate and decreasing the energy consumption.

The major drawback of the third-generation networks is the inefficiency to supply users with new investments. This generation of networks has been characterized as an extension to


second-generation networks. The third-generation partnership project (3GPP) has deficiency in the ability to add new services to the existent ones and satisfy millions of people who use this net-work. Recently, 3GPP has developed new services, such as the merging of two high-demandin g services, multimedia broadcast and multicast service (MBMS) center in combination with an IP multimedia system (IMS). However, the latter did not appeal to users, and these small improve-ments did not encourage customers to change their equipment. Thus, these difficulties have brought communication industries and engineers to the foot of 4G networks, an investment that offers to every user multiple new services with high QoS and at a significantly lower price that could be affordable all over the world. In addition, next-generation networks have to be backward compatible and will have instead a user-centric approach as they will be focused on personalized services [16].

As an example of a cooperative service scenario, imagine a middle-sized area where com-munication through short-range links is possible. This scenario is usual in femtocell networks. Hypothetically speaking, if a number of existing users are trying to access the same service or data from the server, cooperation is possible. First of all, an inquiry through a short-range link is attempted, looking for the desired data in neighboring nodes. If these data exist in one of these nodes, the trade between them can begin, and the destination can get the desired information using a short-range link. Otherwise, if the desired data do not exist in one of the cooperating nodes, then a cooperating download can be applied, that is, users can preagree to download differ-ent parts of the desired information and then exchange them through short-range links. In either scenario, terminals make limited use (or none at all) of the cellular link, using a local retransmis-sions scheme, that way making the reception of data faster and in higher data rates, more accurate, and less energy consuming. In addition, the low use of cellular links greatly facilitates spectrum management as terminals make use of the specific frequencies for less time. Consequently, the spectrum is free most of the time, significantly improving that way the total system capacity. Moreover, short-range communication systems use another part of the spectrum that is not inter-fering with the cellular one. This service is being managed from the lower OSI layers (physical and link layer), making it less complicated and much faster.

Because of the limitations that cellular links have, terminals cannot process high data rates alone. By distributing the downloading, cooperating nodes receive different parts of the file and then, through local retransmissions, every part arrives to the destination node. Enhanced QoS can be achieved that way. An example using multiple description coding (MDC) is demonstrated, but the same framework can be generalized to more types of services. Thus, the bigger the accumula-tion of cooperating users becomes, the faster the data receive to destination and the better quality can be provided to users.

The above services can be seen in implemented applications for cooperative networks. Consider, for example, a bit torrent application for wireless networks that has been implemented on a Symbian OS platform [12]. For the cellular link, GPRS is used, and for the short-range link between terminals, Bluetooth has been preferred. Two terminals want to download the same file from server. Next-generation networks give users the possibility to cooperate and download this file in a cooperative way instead of everyone for itself. Through cooperation, every user downloads a part of the overall file and shares it with the others through Bluetooth. Statistics show that with two terminals the transfer time is half that compared with the noncooperative scenario (in which every user downloads the whole), whereas an energy consumption savings by 44% has been noticed. This behavior lies in the low energy per bit and the higher data rates of Bluetooth [12].


Nowadays, another application that is very popular is video streaming. This application has high functional requirements as it requires extremely low delay and packet loss and binds much of the bandwidth. Additionally, it is very demanding concerning energy consumption, limiting in that way the battery life. In order to fulfill these requirements, some video coding schemes, such as MDC and scalable video coding (SVC) have been introduced. Through MDC, the total stream is divided into small substreams. The quality of the video depends on the number of correct substreams users receive, and different substreams have the ability to be decoded separately. SVC, on the other hand, divides the video into a base layer, which is responsible for the quality of the video, and many enhancement layers. These two techniques can be used in cooperative networks when different users download different substreams from the server and exchange them through short-range links. Whatever type of cellular and short-range links have been used, a significantly high energy and transfer time gain has been noticed [17].

Web browsing is another service that will be reinvented by cooperative networks. The demand for the Internet in mobile phones is growing very fast. Low data rates and high prices have kept this service static for years, disinclining users from using it. 4G networks promise high data rates for web browsing at a significantly low price [18]. Web browsing through cooperation can be extremely fast. There are three phases in the web browsing of a user. The first consists of an inquiry from a user to a server for a specific web page, the second is the download of the necessary data information for this page, and the last phase is the processing of that information. Examining the simple situation where only two terminals exist in a small area (where communication through short-range link, usually Bluetooth, is feasible), the user (master) who wants to download a page makes an inquiry to the second user (slave), so the last one downloads a part of the overall informa-tion from the server. The slave sends the retrieved data to the master through a high-speed short-range link. The capacity of the system improves, and the transfer time decreases significantly that way. In this scenario, it is necessary that the slave terminal does not have any other web browsing activity [18]. The energy, data rate, transfer time, capacity of the system, and spectrum usage gain can be easily visible through statistics and metrics.

All of the above applications and services can be generalized to support more than two termi-nals, improving the overall gain in all the above sections. As a conclusion, we can say that coopera-tive networks can offer great benefits to users and communication industries by introducing new services or by using already existing ones in a more effective way.

In the next scenario, the simple situation of a file download is being examined. Every terminal can support two air interfaces: the cellular and the short-range links. The first one is being used for the communication with the BS, and the second one is used for the communication between terminals. For the first air interface, UMTS communication protocol is being used while, for the second one, terminals use Bluetooth protocol. It is supposed that these interfaces can be used by the terminals simultaneously, meaning that every terminal can receive data from the BS while sending or receiving data through Bluetooth. As it can be seen in the example given in Figure 19.4, MT1 downloads a medium-size file from the BS, while in parallel, it simultaneously can use ter-minals MT2 and MT3 in order to download the file faster.

In a situation of two neighboring terminals, if the first one wants to download the file, then it makes an inquiry (as the master terminal) to the other terminal (slave) in order to cooperate during the downloading phase. With the assumption that the two terminals are located in the Bluetooth coverage area, three different scenarios are examined in the frame of this study. In the first one, it is assumed that the slave has free resources; in the second one, both terminals download the same file simultaneously, and in the third one, the slave does not have free resources.


In the first situation, the slave decides if it will cooperate and give its resources to the master. In the second one, terminals decide either to cooperate (and thus form a cluster, where no distinc-tion between master and slave exists) or not to cooperate (and thus download the whole file by themselves). In both the above situations, if cooperation succeeds, then users can enjoy the benefits mentioned before. In fact, every terminal downloads part of the file and simultaneously sends it to the other members of the cluster [14]. In the last situation, the master terminal can neither use the resources of the slave or form a cluster; thus, it downloads the whole file through a cellular link. In the case that the slave transitions to a pending situation, it can give its resources to the master terminal so as to cooperate [13]. At this point, it must be noted that, as it has been explained in the above scenarios, the basic part of cooperation is simply the existence of many terminals in an area forming clusters that can cooperate in a dynamic way by choosing a master and slaves. This procedure can take place particularly in overcrowded areas where short-range links are feasible.

19.3.3 Limitations and Problems of Cooperative ServicesIn order to render possible the implementation of cooperative networks, specific problems must be solved, and at the same time, motives should be given to users in order for them to cooperate. First of all, in the above-mentioned technologies, the cooperation of intermediary nodes in certain applications (network coding, promotion of packages and data, etc.) was taken for granted; how-ever, this cannot be ensured in real-life networks and applications. More concretely, in many cases, the intermediary node (mobile terminal relay) does not acquire direct profits from cooperation, but only wastes the energy of its battery or even delays the dispatch of its own data if it is already an intermediary node in a call or transport of data. Furthermore, it may even be impossible for the relay to receive packets because all the channels in which it “listens” are occupied.

Another problem that concerns the standardization of cooperative networks is the complex-ity of algorithms that should be developed. Certain algorithms have already been proposed for various fields, such as the finding of the nearest node and the recognition of the topology of the

Base station

Cellular links

Short-range links

MT3

MT2

MT1

figure 19.4 example of three cooperating mobile terminals (Mts).


networks for network coding [19]. However, it is also essential to create algorithms and protocols in the mobile terminals, so they can collaborate efficiently. All of the above increase the complexity needed to design the 4G networks and the corresponding mobile terminals.

Another problem to be solved is the extension of the existing communication protocols and the access techniques so as to cover all different cooperative scenarios. As an example, at Bluetooth, only one master and seven slaves can exist at the same time in each communication. Moreover, slaves cannot communicate with each other; thus, they cannot meet the specifications for coopera-tive networks [12].

Practically, these three problems reveal the existing complexity and difficulty in order to make all the different networks communicate with each other simultaneously. On the contrary, how-ever, despite these complex problems, the results from the collaboration of networks can lead to enormous growth of the branch of networks and communications with all the benefits discussed so far.

19.3.4 Motivations for CooperationAccording to the above, it is essential to provide certain motivations to the intermediary nodes so that they grant their resources to other terminals on demand. A lot of proposals have been made about what these motivations should be. The basic proposal, which is also examined below, is the creation of a mechanism to memorize the nodes that cooperate and the ones that do not.

This mechanism is based on the “reputation” of each terminal, namely, the degree of help that it offers (reputation-based mechanism) [20]. According to this mechanism, the possibility of interaction with each terminal depends on the reputation that it has. The reputation of each terminal can be managed in two ways: centrally or distributed. In a central reputation system, the institution that checks the reputation of the terminals collects elements for them and pre-sents them at the network. Therefore, each user is given the possibility of accessing information relative to the reputation of the other users. In a distributed system, each user stores informa-tion on the neighboring terminals. He or she has therefore his or her own database, and he or she compares it with the other users. The basic idea of this particular mechanism is that as it concerns users that collaborate and have a good reputation, it must give them the possibility of acquiring certain profits [21]. As an example, if one user helped another in the transmission of data acting as a relay, it should be possible in the future to use the other user’s resources for its own profit [13,22].

A second proposal is the use of concrete and more complicated algorithms, aiming at the local-ization of users, both of those who collaborate and those who do not, and taking corresponding decisions [13,21–23]. Moreover, there are also other mechanisms for the control and the benefit of motivations in the synergic networks, which are less popular, such as the mechanism of wage (remuneration mechanism). It should be stressed that these mechanisms import overhead into the system as they require mechanisms of coding for the communication between nodes. Accordingly, the traffic is increased and the capacity of the network is decreased, leading to a degradation of throughput [21,24].

Furthermore, certain motives independent from the mechanism can be used for cooperation, which can be given from each provider of mobile telephony. More specifically, the latter can pro-vide lower debits or even the benefit of better services as higher data rates for the “good” users, that is, for the users that cooperate and provide their resources to support other users or, otherwise, for users with good reputations.


19.4 Simulations and discussion of Cooperative network Scenarios

19.4.1 Simulation ScenariosIn this paragraph, we describe how the aforementioned cooperative framework has been simu-lated. In order to quantify the effectiveness of 4G cooperative networks, the following three sce-narios have been modeled:

◾ Scenario 1 (baseline scenario): First, a UMTS network that serves some users that make use of the file transfer protocol (FTP) service in order to download a file of variable size has been simulated. In that situation, cooperation does not take place, and every user operates autonomously in order to download the specific file from the nearest BS.

◾ Scenario 2 (remote node assistance scenario): In the second scenario, cooperation is supported by introducing Bluetooth links between a user (master) and all the others that are within the radius of Bluetooth coverage, that is, in a distance less than or equal to 10 m. This simulation corresponds to the ability to provide services at a terminal that is not able to communicate with the BS. Thus, nearby nodes or terminals offer their help by providing their resources to the master terminal in order to download parts of the desired files from the BS. These nodes receive data from the BS through cellular links and forward them to the master terminal via Bluetooth. The rest of the parts of the file are downloaded via a cellular link (UMTS).

◾ Scenario 3 (cooperative broadcast scenario): In the last simulated scenario, the effectiveness of cooperative networks is considered. A single user’s mobile device that happens to achieve better data rates from its neighboring nodes through the UMTS protocol becomes auto-matically the master of the cluster and installs Bluetooth links with terminals within the radius of Bluetooth coverage. The master terminal uses its resources in order to help other users download the desired file faster and with significantly lower power consumption.

The multiple variables of the simulation give the ability for an effective examination of the above scenarios in order to obtain the desired results. Some of these parameters are presented below.

◾ Number of users: The maximum number of users that can participate is limited by the Bluetooth protocol that is used in short-range links to a total of eight mobile devices, namely, one master and seven slaves. The number of participating users plays an important role in cooperative networks and especially in scenarios 2 and 3, as this number significantly affects the results.

◾ Dimension of topology: The BS for the cellular network and the mobile terminals have been placed in a three-dimensional coordinate system covering an area of 3375 m3 at random positions. Bluetooth coverage is 10 m, whereas the BS is able to communicate with every terminal independently of their distance.

◾ File size: The file size in our study varies from 10 kB to 1 MB.

19.4.2 Network ModelThe main file of the simulation is divided into IP packets and protocol data units (PDUs). The number of these packets depends on the file size that has been defined for the FTP service and the


size of the IP and PDU packets. In the frame of this work, the IP packet size is equal to 1500 bytes (40 bytes header and 1460 bytes data). The PDU unit size is defined as being equal to 40 bytes, as it has been determined in [25], and a 2-byte header that can be considered negligible.

In a situation of cooperative networks, Bluetooth packets are also transmitted. In that case, the PDUs that master or slave terminals receive through cellular links are encapsulated into Bluetooth packets. Bluetooth packet size varies in every network. In our simulation, considering a low BER (10−4), DH5 packets have been used [26]. Their size is estimated as 339 bytes per packet (15–16 bytes header) and the Bluetooth data rate at 723 kbps [26]. The creation of data packets is different every time, depending on the simulated scenario. In the case of the baseline scenario, the whole file is split into IP packets and then into PDUs. In the case of the remote node assistance scenario, after the inquiry for cooperating users, every cooperating user assumes to download 10% of the whole file. Every time the user downloads a certain number of PDUs, the latter are encapsulated into Bluetooth packets and sent to the master terminal. In the last scenario, every time the master termi-nal receives some PDUs from the BS, it encapsulates them into Bluetooth packets and sends them to cooperating users. More specifically, every Bluetooth packet consists of eight or fewer PDUs.

Data rates and BEP are considered constant at every link. Data rates have a peak of 384 kbps for the cellular link (UMTS) and 723 kbps for Bluetooth. The above data rates are the maximum feasible rates for every link; however, they are rarely achievable for many reasons. BEP highly affects the data rates as errors in transmission lead to retransmitted packets, that is, lower through-put. Consequently, the higher the BEP, the lower the QoS experienced by end users. The time and manner of retransmission is determined by the TCP. The BEP is set to 10−3 for the cellular links and 10−4 for the short-range links.

Furthermore, the TCP, which is the most popular protocol for the transport layer, is respon-sible for the installation of a point-to-point reliable link over the nonreliable IP. This protocol has mechanisms for the control of data flow in order to avoid congestion and has also mechanisms for the packets’ retransmission in order to achieve reliability [27]. In the current study, we do not examine the initial condition for the link installation (such as triple handset); on the contrary, we focus on retransmission mechanisms. The BS, after sending PDUs to users, is waiting for the corresponding ACKs (packets for acknowledgment) from the user. In a situation of negative ACK (NACK) or of a lack of the corresponding ACK, the BS has to retransmit the file. In every trans-mission, a copy of the sent packet goes into a unique buffer (transmission buffer). In case of ACK, the copy of the packet is deleted from this buffer, whereas in case of NACK, it is moved to the retransmission buffer in order to be dispatched again [27,28].

Another thing that should be taken into account is the flow control in which the sliding window mechanism has been used. With this method, the simultaneous dispatch of more than one packet can be achieved, highly increasing the spectrum usage. The packet transmission starts with a small-size window, one to two PDUs, but when BS receives an ACK, this window becomes larger by one until a maximum threshold, which is defined by the bandwidth delay product (BDP), is being reached. In this situation, the window remains stable unless BS receives a NACK, turning the sliding window into half of the current one. For UMTS, BDP varies from 7200 to 64,000 bytes [28,29]. Considered the minimum BDP, BS can send five full-size IP packets, that is, 180 PDUs. This fact could limit the link capacity, however, because of retransmitted packets; this limit is rarely achieved.

Finally, for UMTS, over the physical layer, there is an additional layer, which is divided into the radio link control (RLC) and medium access control (MAC) layers. The RLC layer is respon-sible for segmentation and reassembly of packets to the lower and upper layers, whereas MAC is responsible for the medium access [28,30]. The timeout during the forwarding of packets from MAC to the physical layer, known as TTI, is usually set at 10 ms.


19.4.3 Energy Consumption EstimationFor the study of energy consumption during the simulation, the energy that is consumed by the mobile terminals for their operation must be calculated. Below, we present the model that we use for the assessment of energy both in cellular and short-range Bluetooth links.

19.4.3.1 Cellular Link (UMTS)

The cellular link is the main link for the communication with the BS. Because of the distance between the terminal and the BS, a large amount of energy is consumed from the MT. The model of a cellular system that could calculate with precision the energy that is consumed at this type of connection is difficult to develop. Practically, the biggest percentage of energy for the file download from the BS concerns the time that the mobile remains in the situation of high state, in which it remains so that it can receive all the PDUs in order to reconstruct the entire file [31]. A constant rate of the consuming energy per kilobit can be taken as a result of simulations, presented in the bibliography, namely, 35,121 J/MB or 0.0044 J/kb [32]. In the framework of this study, we estimate, thus, the consumed energy per PDU as 0.0014 J/PDU.

19.4.3.2 Short-Range Link (Bluetooth)

The short-range link (Bluetooth) concerns the communication between MTs. For the purposes of the simulation, it is assumed that the longest distance that the Bluetooth protocol can uninter-ruptedly run between MTs is 10 m. Moreover, in order to create a better measurement model of the energy consumption of MTs, we estimated separately the energy consumed by the master and slave MTs. Based on the literature, the energy that is consumed by the mobile terminals per kilobit and per category (master or slave) can be calculated as follows [33]:

◾ From 0 to 5 m. A stability is observed concerning the distance, so constant values of con-sumed energy are considered:

Master: energy = 0.001 J/kb = 0.0029 J/Bluetooth Packet

Slave: energy = 0.0008 J/kb = 0.0023 J/Bluetooth Packet

◾ From 5 to 10 m. A linear dependence of energy from the distance is observed. Moreover, the bents of interrelations for master and slave MTs can be considered similar [33]. Energy per Bluetooth packets is estimated thus as follows:

Master: energy = 5.74 × 10−5 × d[m] + 2.6 × 10−3, for 5 ≤ d ≤ 10

Slave: energy = 5.74 × 10−5 × d[m] + 2.10−3, for 5 ≤ d ≤ 10

19.4.4 Simulation ResultsIn this section, the results from the extensive simulations of the different scenarios are presented in order to show the possible profits from the future cooperative networks. These results are based


on measurements that took place during a large series of simulations. The analysis of the results can be divided into two parts: the first part concerns the comparison of the two first scenarios, whereas in the second part, the comparison is between the first and the third scenarios. Data rate, transfer time, and the energy consumption are metrics (figures of merit) that we consider in order to perform the comparisons in this study.

19.4.4.1 Comparison between Remote Node Assistance and Baseline Scenarios

It is expected that as more users contribute in downloading, the greater the gain we have in data rate and transfer time for the master terminal. For this scenario, the file size is 500 kB. As it can be seen in Figure 19.5, the master terminal improves its data rate proportionally to the number of cooperating users. The number of cooperating terminals can be less than eight as Bluetooth pro-tocol dictates. In the case of seven contributing nodes, the data rate gain can be more than 100%. The rest of the users do not enjoy any benefit from this cooperation. For these users, as explained before, some motivations have to be given from the service provider in order to cooperate. In addition, transfer delivery time has a remarkable decrease. With cooperation, the master terminal receives parts of the file from other nodes decreasing this way with the transfer time.

Concerning the energy consumption as it is explained in Section 19.4.3, it would be expected to be less than in a single UMTS network, and especially for the master terminal because it receives part from the whole file via Bluetooth links. However, the master terminal wastes 9%–11% more energy compared with a UMTS network. This paradox lies in the fact that the master terminal wastes more energy than slaves to download the file from short-range links. Nevertheless, the overall energy consumption is similar to a UMTS network. Results are depicted in Figure 19.6. As a conclusion, we could claim that, in this specific scenario, we could achieve better data rates and transfer time, wasting the same energy in order to help the master terminal to download the desired file.

Except for the number of cooperating users, the aforementioned metrics depend also on the file size that the master terminal expects to download. In this study, the file size varies from 100 kB to 1 MB, and the number of cooperating users takes discrete values. It is important to mention that the bigger the file is, the greater the data rates are expected, as some time is needed

450400350300250200150100

500

1 2 3 4Number of cooperating terminals

Average data rate—remote node assistance scenario

5 6 7

Scenario 1Remote node assistance scenario (master terminal)

Remote node assistance scenario (all users)

Aver

age d

ata r

ate (

Kbps

)

figure 19.5 average data rate versus number of cooperating terminals (scenario 2).


to reach a peak. More specifically, data rates have a good improvement analogically to the file size. The data rate gain reaches 35%–40%. Generally, the data rate has a stable greater value compared to a UMTS network, and this fact affects greatly the transfer delivery time, which is reduced. Last, but not least, as it is expected, higher file sizes lead to higher energy consumption from terminals.

19.4.4.2 Comparison between Cooperative Broadcast and Baseline Scenarios

In the third simulated scenario, one user who has a better cellular link from his or her neighboring nodes receives inquiries in order to help the transmission and installs Bluetooth links with them. The master terminal downloads parts of the whole file and sends them to other nodes that want this file. Imagine a very popular file, such as a video, that many users want to download in a small area. The transfer of Bluetooth packets does not stop until the whole file is received from users, and this transfer is dynamic. Assuming, however, that all users start simultaneously to use the FTP service, the percentage of the file they receive via Bluetooth is 35%–40%.

In that case, as in the previous scenario, a file of 500 kB is used for the FTP service, whereas the number of cooperating users is variable from one to seven. As depicted in Figure 19.7, the cooperating users have gains bigger than 120% compared to the UMTS network. This gain is independent from the number of cooperating users as the master terminal distributes bandwidth evenly at the slaves. In that way, all cooperating users have the same gain.

The huge gain in data rate leads to an equally great gain in transfer time. This gain is con-stant and does not depend on the number of cooperating terminals because the gain in data rate is stable, and the transfer time is reduced by approximately 50%–60% compared to the UMTS network (baseline scenario).

Because slaves have greater data rates as a result of cooperation and download their files much faster from noncooperative users, there is a great reduction in energy consumption. As before, the energy gain is stable and significantly greater compared to a UMTS network. As can be seen in Figure 19.8, the energy benefit is approximately 50%, whereas if we examine the average energy consumption of all users of the system (whether they participate in cooperation or not), the energy gain varies from 0% to 50%. Thus, there is a great energy saving in mobile devices, leading to


Aver

age e

nerg

y con

sum

ptio

n(jo

ules

)

5 6 7

212120201919181817171616

Energy consumption—remote node assistance scenarioScenario 1Remote node assistance scenario (master terminal)

Remote node assistance scenario (all users)

figure 19.6 energy consumption versus number of cooperating terminals (scenario 2).


improvement in battery life and, simultaneously, an equal reduction in the rates of electromag-netic energy that may be harmful.

In addition, data rate, transfer time, and energy consumption have also been examined in rela-tion to the file size. As long as data rate transfer time and energy gain are stable for this scenario, the average gain for cooperative users and all users is examined. As was discussed before, the number of cooperating users does not affect the metrics in this scenario. Lastly, concerning the energy consumption, a growing gain has been noticed that depends on file size and varies from 0% to 50%. For larger files, the energy gain could be even larger. This derives from the fact that the larger the file is, the faster the simulation time increases, and the master terminal sends a greater percentage of the total file via Bluetooth links.

19.4.5 DiscussionBased on the simulation’s results and the figures that have been presented in the previous section, the superiority of an integrated system with cooperation compared to a single mobile network

Average data rate—cooperative broadcast scenario

Aver

age d

ata r

ate (

Kbps

)


5 6 7

450400350300250200150100

500

Scenario 1Cooperative broadcast scenario (cooperating users)

Cooperative broadcast scenario (all users)

figure 19.7 average data rate versus number of cooperating terminals (scenario 3).

Energy consumption—cooperative broadcast scenario

Aver

age e

nerg

y con

sum

ptio

n(jo

ules

)


5 6 7

201816141210

86420

Scenario 1Cooperative broadcast scenario (cooperating users)

Cooperative broadcast scenario (all users)

figure 19.8 energy consumption versus number of cooperating terminals (scenario 3).


without cooperation, such as the UMTS network, is evident. Particularly in the remote node assistance scenario, data rates are doubled, and the download time is reduced by half. This is the scenario in which only one user has benefits. Moreover, the results on data rates are impressive, but they are dependent on the number of cooperative terminals.

In the cooperative broadcast scenario, the energy gain is increased along with the size of the downloaded file. In addition, there is no dependence on the number of cooperative terminals. The only dependence that exists is that of the master terminal. The faster the master terminal down-loads the file, the faster it can seed the parts of the file to the slaves. In other words, the speed of the Bluetooth protocol is constrained by the lower speed of the UMTS protocol. That problem is caused because the master terminal should download the parts of the file before it can seed them. The most encouraging part of this simulation scenario is that as long as the slaves download the most parts of the file from the master, better data rates are observed, approaching the theoretical value of 723 kbps because they reach the average speed of 600–650 kbps. Nevertheless, the gain from this scenario is very important.

As a final remark, simulation results give a sense of optimism that 4G networks could sup-port data rates over 1 Mbps, and their energy consumption needs can be even less if user nodes cooperate. Of course, a lot of obstacles have to be overcome before these simulated scenarios can be deployed in real-life networks, and similar results can be achieved.

19.5 future research directionsOne of the issues that should be taken into account is that the protocol that is used for the second-ary air interface can make a difference. Considering the case of using a Wi-Fi Internet connection, we can get a result of a tremendous change of time and data rate in a positive way. Based on the results of the simulation, it can be ensured that researching the field of cooperation services and networks will not be a waste of time. More research can be done in many cooperative communica-tion fields, but both mobile users and mobile telecommunication providers have to understand the gain of cooperative services.

In addition, the evolution of a mobile terminal’s software that can support cooperative services is one of the most interesting aspects. The adjustment of the protocols that can ensure a communi-cation without fading and interferences between the objects can create a stable environment where different protocols work undisturbed, providing even better data rates and even less energy con-sumption. Because the right trade between mobile users is vital to make users cooperate, it is essen-tial to create the right software, which tracks down the use of cooperative services and, of course, merit the cooperative users. Novel software development theories and supporting technologies need to be introduced, given the IoT software system context awareness and dynamic deployment capability in order to realize the environmental adaptability of IoT services. IoT software should be able to adapt to the dynamic environment flexibly to provide intelligent services.

19.6 ConclusionsThe main goal of this chapter was to discuss the benefits from the development of cooperative net-works. First, an analysis of the IoT was made and the possible benefits from the cooperation strategies on the IoT concept were discussed. In the second section of this chapter, we described the coopera-tive network characteristics, and we presented the benefits from their use. Finally, we presented some


simulation results of a cooperative service, demonstrating the advantages of cooperative networks in energy consumption minimization and the increase in data transfer speed. Results show that many benefits can be derived from node cooperation. Nevertheless, many technical challenges of coopera-tive networks are still open for future research, such as interoperability issues or motivations for coop-eration. Even though theoretical cooperative scenarios can be well formulated and simulated, in real life, incentives have to be provided to rational nodes in order for them to share their valuable resources.

The clearly emerging vision of future networks is that the coexisting heterogeneous wireless networks and mobile terminals will evolve so as to cooperate with each other in order to facili-tate network traffic and guarantee QoS requirements to even the most demanding users. Large-scale applications of IoT for surveillance (e.g., for forests) currently under development integrate wireless-sensor networks with vehicle-carrying networks, intelligent power-usage networks, 3G mobile networks, and the Internet in order to establish multiapplication verification platforms for IoT. Access to large-scale heterogeneous network elements and massive data exchange among them are the important novel features associated with the wide application of IoT. At the same time, network elements in each local region must be able to self-organize dynamically and realize interconnection in an interoperable manner. Consequently, one of the most important challenges related to IoT is the support of data exchange in large-scale heterogeneous network elements with simultaneous local dynamic autonomy.

references 1. Tan, L., and Neng Wang. “Future Internet: The Internet of Things.” 3rd International Conference on

Advanced Computer Theory and Engineering (ICACTE), 2010, 376–380. 2. Puyolle, G. “An Autonomic-Oriented Architecture for the Internet of Things.” Modern Computing,

2006. JVA ’06. IEEE John Vincent Atanasoff 2006 International Symposium, Sofia, 3–6 October 2006, 163–168.

3. Pereira, J. “From Autonomous to Cooperative Distributed Monitoring and Control: Towards the Internet of Smart Things.” at ERCIM Workshop on eMobility, Tampere, 30 May 2008.

4. Takaragi, K., M. Usami, R. Imura, R. Itsuki, and T. Satoh. “An Ultra Small Individual Recognition Security Chip.” IEEE Micro 21, no. 6 (2001): 43–49.

5. Kortuem, G., D. Fitton, F. Kawsar, and V. Sundramoorthy. “Smart Objects as Building Blocks for the Internet of Things.” IEEE Internet Computing, (January/February 2010): 44–51.

6. Juels, A. “RFID Security and Privacy: A Research Survey.” IEEE Journal on Selected Areas in Communications 24, no. 2, (2006).

7. Koshizuka, N., and K. Sakamura. “Ubiquitous ID: Standards for Ubiquitous Computing and the Internet of Things.” Pervasive Computing 9, no. 4, (October–December 2010): 98–101.

8. Atzori, L., A. Iera, and G. Morabito. “The Internet of Things: A survey.” Computer Networks, 31 May 2010. 9. Terziyan, V., O. Kaykova, and D. Zhovtobryukh. “UbiRoad: Semantic Middleware for Cooperative

Traffic Systems and Services.” International Journal on Advances in Intelligent Systems 3, nos. 3 and 4 (2010): 286–302.

10. Baeg, S. H., J. H. Park, J. Koh, K. W. Park, and M. H. Baeg. “Building a Smart Home Environment for Service Robots Based on RFID and Sensor.” Control, Automation and Systems, 2007, ICCAS ’07, International Conference, 17–20 Oct. 2007, 1078–1082.

11. Prasad, N. R. “Secure Cooperative Communication and IoT: Towards Greener Reality.” CTIF Workshop 2010, May 31–June 1, 2010.

12. Zhang, Q., F. H. P. Fitzek, and M. Katz. “Evolution of Heterogeneous Wireless Networks: Towards Cooperative Networks.” Third International Conference of the Center for Information and Communication Technologies (CICT)—Mobile and Wireless Content, Services and Networks—Short-Term and Long-Term Development Trends, Copenhagen, Denmark, November 2006.

Cooperative Services in next-generation wireless networks

Documents