| On Achievable Delay/Capacity Trade-os in Mobile (2004) | |||||||||||||||||
Abstract | |||||||||||||||||
| Recent work of Gupta and Kumar (2000) has shown that in a multi-hop wireless network the throughput capacity per source-destination pair goes to zero as the node density increases. While it has been shown that a constant throughput scaling per source-destination pair can be achieved in mobile ad hoc networks, the delay related aspects have not been considered in detail. In this paper, we study the delay-capacity trade-o# in mobile ad hoc networks. We consider two canonical random mobility models in this paper; the Brownian mobility model (BMM) and the random way-point mobility model (RWMM). We show that under the distributed 2-hop relaying protocol proposed by Grossglauser and Tse (2001), the packet delay scales as #(T p (n)n) under the RWMM and O(T p (n)n log (n)) under the BMM, where T p (n) is the packet transmission time. We then show that the delay scales as (T p (n) # n), under a broad class of scheduling and relaying protocols. Further, we show that the trade-o#: delay/capacity #(T p (n)n), is necessary as well as su#cient under our settings. We then propose two distributed protocols which achieve the above mentioned lower bound on the packet delay, and evaluate their performance in terms of the delaycapacity trade-o#. | |||||||||||||||||
Publication details | |||||||||||||||||
| |||||||||||||||||