Algorithms for opportunistic load balancing cognitive engine

This paper presents a cognitive engine (CE) that intelligently and dynamically allocates spectrum resources to users in the following two cases: 1) a scenario where a network has an exclusive access to a spectrum band; and 2) a spectrum sharing scenario where along with the exclusive utilization to its own spectrum band a network also can opportunistically utilize shared spectrum band as a second user. Moreover, the implemented CE performs two main tasks: 1) Accepts or rejects arrival user requests based on a priority based algorithm; and 2) It intelligently load balances the user traffic between the two available network resources, while taking into account the primary user activity in the shared spectrum band. We evaluate performance of the proposed algorithms under different primary and secondary user traffic scenarios. We show that the proposed load balancing algorithm increases average throughput of the network and it also reduces the average number of users rejected by the network.