OLSR Control Overhead and Power Consumption Reduction using Game Theory

Abstract

Recently, MANET has been gaining the popularity due to its ease of implementation. One of the major elements in MANET is routing protocol, which consists of two main protocols: proactive and reactive routing protocols.

In this work, we focus on proactive protocol called OLSR (Optimize Link State Routing Protocol), where the routes are always maintained by interchanging control overhead, namely HELLO and TC (Topology Control) messages. However, this is very wasteful of resource and causing the performance degradation. We propose the method to reduce the overhead while maintain the throughput of OLSR and also reduce the power consumption by using the well-known mathematic tool, widely used in interactive decision systems, called game theory. Our proposed method is called Game Theoretical OLSR (gOLSR). We also investigate the effect of gOLSR based on another type of Medium Access Control (MAC) protocol called Sensor Medium Access Control (SMAC) protocol used in mobile sensor network.

OLSR is modified in such a way that every node in the system has to play the game when HELLO and TC interval are expired. Each node will choose its strategy to “Update” or “Not Update” the HELLO and TC messages in each round of game. The performance in terms of control overhead and throughput of the proposed algorithm is evaluated by using parameters namely Normalized Routing Overhead (NRO), Average Throughput and Normalized Overhead Reduction Index (NORI). However, the performance in terms of power consumption is evaluated by metrics called Average Power Consumption of nodes in two states (transmission and reception).

According to the simulation results, it is apparent that the proposed algorithm provides large NRO reduction while the Average Throughput is reduced a little bit. The power consumption of the network in all states is also reduced. That is, the proposed algorithm can reduce the certain amount of Control Overhead as well as the Power Consumption while the Average Throughput is reduced a little bit. Therefore, Game Theoretical OLSR (gOLSR) is able to support energy efficiency MANET in various node mobility and node density environments.