Abstract: In structured P2P networks, the skewness in query pattern can result in load imbalance and even congestion. To address the problem, this paper presents a dynamic congestion control algorithm based on indegree adjustment approach. The basic idea behind the algorithm is that it dynamically identifies a group of nodes near a hot spot and then reallocates loads among them by indegree migration. The results of the simulation experiments show that the algorithm can effectively reduce the level of congestion in case of load burst with a low overhead.

Key words: P2P networks, congestion control, load balance, hot spot, indegree adjustment

摘要： 针对结构化对等网络中存在的由查询负载不均衡所导致的拥塞问题，提出一种基于动态入度调整的拥塞控制算法。算法根据节点查询负载的情况，动态判定出可能存在的热点，并将热点周围的节点组织起来，通过调整每个节点入连接的数量对负载进行调整。为了克服节点处理能力的异构性，算法将热点组中的节点进一步划分为多个能力组，并在能力组内和组间分别进行调整。模拟结果表明，该算法在处理突发流量的过程中可以有效地减少查询失败的次数，降低拥塞产生的影响。

关键词: 对等网络, 拥塞控制, 负载均衡, 热点, 入度调整

CLC Number: 

• TP393