Interference Aware Routing Using Localized Mobility Prediction for Multihomed Wireless Networks

Abstract — In this paper, we present two novel algorithms to deal with mobility prediction and interference aware routing for multihomed wireless networks. First, a localized mobility pre-diction algorithm, LMP, is developed using the Hidden Markov Model (HMM) in which the multiple fixed relay nodes in the multihomed network architecture act as pseudo-base stations to locally maintain and deliver mobility information collected from surrounding mobile users. We show that the prediction accuracy of our proposed prediction algorithm is better than using common Markov chains to predict user location at a time instant t. We also show that our mobility prediction algorithm adapts better to a user node’s change in movement. Second, we present a new interference aware routing algorithm in which the signal to interference noise ratio (SINR) is used as the routing metric to determine least interfering paths. The mobility prediction algorithm is used as input to the routing algorithm in order to accurately calculate the SINR value of a specific link at particular time instances. This information is used to perform route construction based on least interference. We solve the least interference routing problem using a minimal cost flow optimization framework. We show that the integration of the two algorithms outperforms conventional counterparts in the literature in terms of packet delivery ratio and end-to-end-delay. However, we also show that the tradeoff for increased network performance lies in the ability of the algorithm to scale to very large networks