Local information-based congestion control scheme for space delay/disruption tolerant networks

Abstract

The storage resources and communication opportunities in space delay/disruption tolerant networks (DTN) are usually very limited. Moreover, as the links in space DTN are often subject to long delay, intermittent connectivity and asymmetric bandwidth, end-to-end continuous path may not be guaranteed and messages may stay at intermediate nodes for a long time. Thus, congestion in space DTN, which takes the form of persistent storage exhaustion, is inevitable. Due to the intrinsic features of space DTN, congestion control should be performed with local information before congestion occurs and limited storage and communication resources should be allocated to messages that have the highest probability to be delivered to destination. Based on these principles, a local information-based congestion control (LCC) scheme for space DTN is proposed in this paper. Firstly, LCC attempts to relieve the storage pressure by using alternative paths i.e. non-best paths when the network is about to congest. Secondly, LCC adopts a utility function based on the reciprocal of hop counts to destination to assist forwarding decision and queue management. Simulation results show that with the combination of LCC typical space DTN routing algorithm can achieve higher message delivery ratio and more uniform traffic distribution.