Using incompletely cooperative game theory in wireless sensor networks

Liqiang Zhao, Hailin Zhang, and Jie Zhang

WCNC 2008Presented by: Abolfazl Asudeh

Using Incompletely Cooperative Game Theory in Wireless Sensor Networks

04/09/20232

Outline Background and Problem statement Incompletely cooperative game GMAC and Simulation Results

04/09/20233

Wireless Sensor Networks (WSNs) have a wide range of potential applications including environment monitoring, smart spaces, medical systems and robotic exploration.

Energy Consumption is usually the more important metric other metrics: Delay, Throughput, …

A lot of interest in performance analysis and improvement, specially in Medium Access Control (MAC) protocols.

04/09/20234

Main Categories of MAC protocols of WSNs Scheduled Protocols: use a predefined

schedule for data transmission - TSMP

Protocols with Common Active Period: Nodes wake up and sleep together to reduce the energy cost of idle listening. - SMAC idle listening: when the nodes are awake, but

there is no packet in the channel to be transmitte Usually use CSMA/CA (RTS/CTS) for data

transmission

Receiver

Tg

Data Tg

Ack

Sender Data Tg

Ack

Time Slot

04/09/20235

Main Categories of MAC protocols of WSNs Preamble Sampling Protocols: Every node

wakes up periodically and check if there is a packet in the channel; the sender sends a long preamble before sending the data. – Preamble Sampling Aloha

Hybrid Protocols: the combinations of the protocols of different categories

Sender

Receiver

Preamble Data

Data Check Interval

04/09/20236

Game Theory is used broadly in distributed WSNs to solve problems like: Security Routing Power Control

The goal of this paper: to achieve energy consumption, delay, and throughput together

Main focus is on Common Active MAC protocols

04/09/20237

Sensor networks are energy constraint explicit cooperation among nodes is impractical.

It can be modeled as an incompletely cooperative game, a stochastic game, which starts when a new packet arrives at the node’s transmission buffer and ends when the packet is transmitted successfully or discarded.

Each game process includes many timeslots and each timeslot corresponds to one game state.

04/09/20238

In each timeslot, each player (i.e., node) estimates the current game state based on what happened in the past timeslots.

Then adjusts its own equilibrium strategy by tuning its local contention parameters

04/09/20239

Framework Contains three major components

a detector: detect and record current conditions an estimator: estimates current state, such as

number of nodes an adjustor: makes the decision for the strategy

and adjusts the contention parameters

04/09/202310

The Game model of N+1 nodes

04/09/202311

Estimator: Estimates the current state

Adjustor: changes the transmission probability by tuning the contention parameters: CWmin

=[n*rand(7,8)] It is shown that the ratio of optimal CWmin to n is

almost from 7 to 8 Collision Window

Random Back-off

CWmax

04/09/202312

GMAC: the simplified MAC solution The Estimation of n is inaccurate in

unsaturated environments.

GMAC is the simplified version.

In G-MAC, after transmitting a packet, no matter it is transmitted successfully or not, the player does not start the next game process with the nominal CWmin.

04/09/202313

Simulation Results channel rate: 1Mbps n = 30 packet arrival: Poisson active period: 250msec

04/09/202314

References A. Bachir, M. Dohler, Th. Watteyne, and K.K.

Leung, “MAC Essentials for Wireless Sensor Networks”, IEEE COMMUNICATIONS SURVEYS & TUTORIALS, VOL. 12, NO. 2, SECOND QUARTER 2010.

Liqiang Zhao; Hailin Zhang; Jie Zhang; , "Using Incompletely Cooperative Game Theory in Wireless Sensor Networks," Wireless Communications and Networking Conference, 2008. WCNC 2008. IEEE , vol., no., pp.1483-1488, March 31 2008-April 3 2008

04/09/202315

Thank you

Question?