A serve flow management strategy for IEEE 802.16 BWA system in TDD mode

A serve flow management strategy for IEEE 802.16 BWA system in TDD mode

Hsin-Hsien Liu

2005 11 15

Outline

Introduction PHY and MAC layers The QoS management for IEEE 802.16 Proposed service flow management Simulation results Conclusion

Introduction

IEEE 802.16 defines the air interface and MAC protocol for a WMAN, intended for providing high-bandwidth wireless voice and data for residential and enterprise use

The first version was completed in December 2001 10-66 GHZ, 32-134Mbps

802.16a was completed in January 2003 2-11 GHZ, up to 75 Mbps

Introduction

802.16d upgrade to the 802.16a was approved in June 2004 (now named 802.16-2004) and primarily introduces some performance enhancement features in uplink

802.16e is underway, which to support mobility up to speeds of 70-80 mi/h

Their main advantage is their fast deployment which can result in cost savings

Introduction IEEE 802.16 MAC protocols have been proposed

to support QoS guarantees for various kinds of applications

IEEE 802.16 left the QoS based packet-scheduling algorithms that determine the uplink and downlink bandwidth allocation, undefined

Introduction Several approaches for bandwidth allocation for

TDD mode, they only consider the scheduling for uplink sub-frame

Since most paper applies strict priority queue for different class of service, which leads starvation of low priority service when higher priority service is heavy

PHY and MAC layers

The basic architecture consists of one Base Station (BS), and one or more Subscriber Stations (SSs)

The BS regulates all the communication in the network

PHY and MAC layers

The communication path between SS and BS has two directions Downlink channel (from BS to SS) Uplink channel (from SS to BS)

IEEE 802.16 has been designed to support FDD and TDD

PHY and MAC layers

On the downlink, the data packets are broadcasted to all SSs and an SS only picks up the packets destined to it

On the uplink, the BS determines the number of time slots that each SS will be allowed to transmit in an uplink subframe

Uplink map message (UL-MAP) contains information element (IE), which include the transmission opportunities

PHY and MAC layers

After receiving the UL-MAP message, the stations transmit their data in pre-defined time slots as indicated in the IE

A scheduling module for the UL is necessary to be kept in the BS in order to determine the transmission opportunities using the bandwidth requests sent by the SSs

Service flow Unsolicited Grant Service (UGS)

Support real-time service flows that generate fixed-size data packets on a periodic basis

It allocates a fixed numbers of time slots in each time frame

Real-Time Polling Service (rtPS) Support real-time service flows that generate variable si

ze data packets on a periodic basis

Service flow

Non-Real-Time Polling Service (nrtPS) Support delay-tolerant data streams consisting of vari

able-sized data packets for which a minimum data rate is required

Best Effort Service (BE) Support data streams for which no minimum service l

evel is required and therefore may be handled on a space-available basis

The QoS management for IEEE 802.16 Admission control

It is used to limit the number of flows admitted into the network

Buffer management It is deployed to control the buffer size and decide whic

h packet will drop Scheduling

It is adopted to determine which packet will be service first in specific queue to guarantee its QoS requirement

The QoS management for IEEE 802.16 Since 802.16 MAC protocol is connection

oriented, the application must establish the connection with BS as well as the associated service flow

BS will assign the connection with a unique connection ID (CID) to each uplink or downlink transmission

When a new service generates or updates its parameters, it will sent message (DSA/DSC) to the BS

The QoS management for IEEE 802.16

Proposed service flow management for IEEE 802.16 The hierarchical structure of bandwidth

allocation

Proposed service flow management for IEEE 802.16 Bandwidth requirement can be measured by the

maximum sustained traffic rate (rmax) and the minimum reserved traffic rate (rmin)

rmax and rmin are carried in the DSA and DSC message at the beginning period of connection setup

The minimum reserved traffic rate is used for admission control

The maximum sustained traffic rate is used for scheduling

Admission control One principle is to ensure the exiting connection’s

QoS will not be degraded significantly and new connection’s QoS will be satisfied

For those connections whose Minimum Reserved traffic rate is equal to zero, they can always be accepted, but the QoS will not be guaranteed

Scheduling architecture First layer scheduling: Deficit Fair Priority Queue

(DFPQ) There is an active list maintained in BS The DFPQ only schedules the bandwidth application

services in the active list If the queue is empty, it will be removed from active

list The service flows in active list are queued by strict

priority shown in Table 1

Scheduling architecture

First layer scheduling: DFPQ The scheduler visits each non-empty queue in the

active list and determines the number of request in this queue

The variable Deficit Counter is incremented by the value Quantum each time when it is visited

Scheduling architecture First layer scheduling: DFPQ

If Deficit Counter is more than zero , the variable Deficit Counter is reduced by number of bits in the packet and the packet is transmitted to the output port

The process will be repeated until either the Deficit Counter is no more than zero or the queue is empty

If the queue is empty, the value of Deficit Counter is set to zero

When this condition occurs, the scheduler move on to serve the next non-empty priority queue

Scheduling architecture

Second layer scheduling Three different algorithms are assigned to three class

es of service to match its requirement rtPS connection: earliest deadline first (EDF) nrtPS connection: weight fair queue (WFQ) BE connection: the remaining bandwidth is allocated t

o each BE connection by round robin (BB)

Scheduling architecture

Buffer management Used to control the buffer size and decide which

packets to drop Timing sensitive traffic has its maximum delay

requirement Buffer management will drop those packets that

exceed their maximum delay

Simulation results

The assumption of total bandwidth is 10Mbps The duration for each frame is 10 ms, so the ban

dwidth for a frame is 100Kbit All packet arrivals occur at the beginning of each

frame and the packet arrival process for each connection follows the Poisson distribution with different traffic rate λ

Simulation results

Simulation results

Conclusion

A 2-layer service flow management architecture for IEEE 802.16 is proposed

Compared with fixed bandwidth allocation, the proposed solution improves the performance of throughput under unbalanced uplink and downlink traffic

Better performance in fairness can be achieved by the proposed DEFQ algorithm than strict PQ scheduling

Reference IEEE 802.16 Standard-Local and Metropolitan Area Net

works-part 16. IEEE 802.16-2004 Jianfeng Chen; Wenhua Jiao; Hongxi Wang; “A Service f

low Management Strategy for IEEE 802.16 Broadband Wireless Access Systems in TDD Mode”, Communications, 2005. ICC 2005. 2005 IEEE International Conference onVolume 5,  16-20 May 2005 Page(s):3422 - 3426

K. Wongthavarawat, and A. Ganz , “Packet Scheduling for QoS Support in IEEE 802.16 Broadband Wireless Access Systems”, International Journal of Communication Systems, Vol. 16, P81-96, 2003

http://www.google.com/