In resource-limited wireless sensor networks(WSNs), network resource will run out soon due to communication of mass redundant data. We employ Network Coding as an effective in-network processing strategy to reduce the data transmission amount. With this concern, a comprehensive Network-Calculus based analytical framework for Cluster-tree WSNs involving network coding is established in this paper. To specify, we firstly derive the service curve for sensor node performing network coding; then some useful performance expressions like input/output arrival curve, buffering requirements and end-to-end delay are calculated; finally, to show the effectiveness and superiority of our proposed analytical framework, we demonstrate how to apply the analytical framework to a special IEEE802.15.4/Zigbee protocol. Indeed, numerical results show that network coding does improve buffer requirement and end-to-end delay compared with classical scheduling strategy.