Abstract:
This thesis focuses on network performances and how to solve the problem of low throughput and long delay in both Slotted Aloha and CSMA Medium Access Control (MAC) protocols. For this purpose, we propose two complementary solutions. The first consists of the integration of the erasure coding scheme in this protocol to recover collided packets and to reduce the rate of collision between transmitted packets. Here, since each node sends N coded packets instead of the k original packets, we have (N-k) redundant packets. The introduction of redundancy and subsequently structuring it in an exploitable manner, allows
serious errors injected by the channel to be corrected. However, if each node attempts to achieve its best output without regard for the other nodes’ actions, this could affect overall system throughput. To analyze such conflicting situations where the action of one node has an impact on the other nodes’ actions, we add a complementary solution, which is based on the game theory technique of acquiring network equilibrium. This makes the network stronger and able to resist many collisions. It has been proved that the integration of Erasure Coding in Slotted Aloha and CSMA protocols makes the network more effective to success its transmissions. When, it can correct collided packets and recover all original sent packets.
However, use of random type of erasure coding could increase the average traffic load that leads to network saturation. Therefore, we present a Game model for Slotted Aloha and CSMA networks to study the interactions between users and guide the system to operate in its equilibrium. The simulations demonstrate that at equilibrium the network present a considerable progress in terms of throughput, packet loss ratio and overall delay transmission.
