In-network caching mechanisms for intermittently connected mobile users

Abstract

This paper examines en-route caching of personal content for mobile users with intermittent connectivity. Our goal is to optimize in-network caching in order to reduce the number of retrieval attempts for each content request. Therefore, we design two distributed caching techniques: in last router caching, a router caches a file only when it detects that the end user has disconnected; in price based caching, users tag their files with a maximum acceptable caching price while routers individually establish caching price to ensure that the average cache holding time is longer than the duration of user disconnection. In both approaches, the least recently used (LRU) cache replacement policy is employed and a reconnecting mobile user seeks the closest cached file copy. Using simple analytic models to set caching prices, we compare these techniques against baseline methods that either perform no caching, or cache files at every intermediate router, or make random caching decisions at each router. Simulation results suggest that our proposed methods reduce the number of retrieval attempts of highly mobile users, without penalizing stationary users.