Research On Application Of Differentiated Queueing Service(DQS)in Opportunistic Networks

The opportunistic networks do not rely on the full path for the data transmission. This characteristic leads to its adaptability and vitality in the extremely complex network environment, so it has a broad application prospect. The opportunistic networks adopt the message transmission mode of "storage- carry- forward". However, nodes communication is opportunistic, and the node cache capacity is limited, which will directly affect the performance of message transmission. Therefore, the design of efficient queue scheduling and cache management strategies and the optimization of the message forwarding and discard mechanism will play a key role in improving the performance of the message transmission.Distributed Queuing Service(DQS) is an excellent scheduling algorithm designed to provide better Qo S for different types of data services based on data package units. Its application in traditional networks has greatly improved the network performance. However, there is still no DQS study for opportunistic networks in the opportunistic networks. Therefore, this article attempts to apply the theory of DQS scheduling algorithm into the opportunistic networks, aiming at improving the performance of its message transmission.First of all, regarding the key application of DQS algorithm in the opportunistic networks- the estimation of the remaining delivery delay, this paper presents a method of passive estimation method based on historical information statistics. Through the historical hops and time information carried by network messages themselves, expected hops needed to transfer messages between nodes and the average time required for each hope are estimated. As a result, the remaining delivery delay needed to transfer from the current node to the desired destination node can be estimated.On this basis, the paper proposes a strategy of nodes queue scheduling and cache management based on the estimation of remaining delay. The messages’ scheduling and discard priorities are determined according to the remaining lifetime and estimated remaining delivery delay. So the messages which are more urgent and with higher transfer success probability in the cache queue will be transferred as many as possible within the communication time limit, so that the message delivery success rate can be improved.Finally, the proposed strategy of queue scheduling and cache management based on the remaining delay estimation is applied on the ONE. By changing the node moving rate, node density, cache size, message load and other factors, its performance is analyzed. A comparative analysis is also carried out with the traditional random queue scheduling and cache management strategies. The simulation results show that the strategy can effectively improve the message delivery success rate and reduce the average network overhead, but also increases the average delay of message delivery.