A Resource Reservation Algorithm with Muti-parameters

Advance reservation as an effective technique to support QoS guarantees the availability of resources at specific time. However, it will produce resource fragmentations for processing resource by time sequence, moreover result in lowering resource utilization rate and rising rejection rate in which reservations increased. Based on advance reservation studies, a reservation algorithm with Muti-Parameters called MPRAR for short is proposed in this paper, which means to mitigate these negative effects in our work. We designed a global queue to store reservation requests called FIFO and another queue named Heap which arranged in the order of weight. Weight value is the determinant whether the reservation request would be accepted. The experimental results show that MPRAR can achieve a lower rejection rate and lower mean waiting time compared with the conventional reservation policy and backfilling technique.