In this letter, we propose a distributed jamming strategy to maximize the achievable ergodic secrecy rate (ESR) of a single-input multi-output (SIMO) transmission with a multiple-antenna eavesdropper based on only the statistical channel state information (CSI). In the scheme, exploring the heterogeneous large-scale fading effects, multiple geographically distributed single-antenna jammers transmit independent and uncoordinated jamming signals to confound the eavesdropper. We derive a large-scale asymptotic approximation of the achievable ESR, and optimize the transmit power of each jammer for maximizing the asymptotic ESR by geometric programming (GP). The proposed scheme does not require signal coordination among jammers and greatly reduces the training/synchronization overhead, which is applicable when the receiver has a large number of antennas, i.e., massive MIMO system.