Opportunistic Mobile Data Offloading Using Machine Learning Approach

Currently, cellular networks both 3G and 4G are heavily overloaded due to increasing usage of mobile applications. Offloading mobile data traffic through opportunistic communications is one of promising solution to solve this problem. This has huge advantage over other kinds of offloading techniques like no extra cost, significant reduction of mobile traffic, high efficiency. As a case of study, we are focusing our investigation in opportunistic communication for optimizing target set selection problem. A new algorithm is proposed for generating target set which uses machine learning paradigm. Since, machine learning is an emerging sector in computer science with lots of potential; we integrated it with offloading procedure to achieve better performance in real world scenario. The efficiency of proposed method is measured by comparing it with other methods like Greedy, Heuristic and Random. A case study is conducted incorporating this approach and performance evaluation is done. It can be ensured from this comparison that the proposed algorithm outperforms its counterparts.