Cost-efficient selective network caching in large-area vehicular networks using multi-objective heuristics

In the last decade the interest around network caching techniques has augmented notably for alleviating the ever-growing demand of resources by end users in mobile networks. This gained momentum stems from the fact that even though the overall volume of traffic retrieved from Internet has increased at an exponential pace over the last years, several studies have unveiled that a large fraction of this traffic is usually accessed by multiple end users at nearby locations, i.e. content demands are often local and redundant across terminals close to each other, even in mobility. In this context this manuscript explores the application of multi-objective heuristics to optimally allocate cache profiles over urban scenarios with mobile receivers (e.g. vehicles). To this end we formulate two conflicting objectives: the utility of the cache allocation strategy, which roughly depends on the traffic offloaded from the network and the number of users demanding contents; and its cost, given by an cost per unit of stored data and the rate demanded by the cached profile. Simulations are performed and discussed over a realistic vehicular scenario modeled over the city of Cologne (Germany), from which it is concluded that the proposed heuristic solver excels at finding caching solutions differently balancing the aforementioned objectives.