Applying Bipartite Supergraphs to Mitigate Ping Pong Effect in Hierarchical Wireless Networks

Wireless networks with mesh topology usually have a partitioned, clustered or cellular structure wherein there is a central cluster head/gateway/evolved Node B (eNB) around which a fixed service set is defined. The fragile nature of the wireless network topology and the movement of nodes for load management or for hand off and handover results in instability of the network. The instability happens because, the end node, which can be a User Equipment (UE) or a Wi-Fi client node, gets pushed to and fro between multiple cluster heads/gateways/eNB repeatedly. This happens because the parameter being considered for node transfers and assignments might be at a threshold value for both the sender and receiver node. This is called the ping pong effect. This paper attempts to address this very common issue by applying the concept of Bipartite graph to derive a stability condition. This condition, when applied, on node migrations, will ensure that there is no ping pong effect. The entire computation is performed by using the Supergraph of network which further reduces the computation, thereby rendering the entire scheme very scalable.