This paper investigates the maximum lifetime data gathering problem theoretically. Specifically, (1) the simplified static routing scheme where only one routing tree is used to gather data during the lifetime of network is analyzed, (2) the actual dynamic routing scheme where a series of routing trees are used to gather data is analyzed,(3) a near optimal maximum lifetime data gathering and aggregation algorithm MLDGA is proposed, which tries to minimize the total energy consumption in each round and maximize the lifetime of a routing tree used in the round,(4) the MLDGA algorithm is simulated in Java programming language. Comparing with the existing algorithms that are only efficient in some specified conditions, the simulation results show that MLDGA performs well regardless of base station location and initial battery energy levels of sensors.