Abstract
This research considers the router allocation problem in situations with sensitive, highest-priority, and severe demand constraints. Examples include bespoke communication, masquerade communication, war, disasters, and various crises (political, social, etc.). Specifically, by introducing a network component called the scheduler. The scheduler is established to provide the intelligencers the highest priority to forwarding their packets based on their preferences at any time. Intelligencers define their preferences to control the data release time, due time, and static window transmission time-slot. Each window represents a reserved time-slot for the intelligencer. A window is an interval when the intelligencer can send data without any contention and can block any other user trying to gain access to the router during this interval. The reserved interval is called a "static urgent window pass". For each block of the received data, the scheduler solves a scheduling problem with the windows reserved for the router based on the given constraints, which is an NP-hard problem. This work proposes eight heuristics based on dispatching with constraints, iterative randomization method, and subset-sum method to solve the studied problem. These heuristics are utilized by the two proposed algorithms to forward data from the sender to the receiver based on the specified constraints. The first algorithm prioritizes the set of packets to be sent and is called the "priority packet-based algorithm". The second algorithm is called the "cumulative packet-based algorithm" and is based on the non-priority set of the remaining packets. At any time other than the specified window, the scheduler rearranges and reschedules all the remaining (non-scheduled) packets. Experimental results showed the performance of the proposed heuristics in terms of gap and time. In the performed experiments, the scheduler was capable to pass the router role and gain the control to send the required data based on the specified constraints.