Dynamic bandwidth provisioning using Markov chain based RSVP for unmanned ground networks

An unmanned ground vehicle (UGV) network operates as an intermittently connected system, distributed over large geographic areas. The cooperative nature of a UGV network requires that bandwidth be allocated depending on necessity between individual UGV nodes. In this paper, we study the problem of dynamic bandwidth provisioning in a UGV network. Specifically, we integrate the use of a basic statistical model, known as the Markov chain with a widely known network reservation protocol known as the Resource Reservation Protocol (RSVP). The Markov chain is used to estimate future traffic demand along a path based on channel conditions. The Markov chain results are used as input to the RSVP which then allocates the required bandwidth along the path. Using a wireless network simulation platform called Qualnet, we show that the integrated Markov chain and RSVP algorithm provides higher bandwidth guarantees and better overall quality of service (QoS) when compared solely with using RSVP in wireless communication networks.