Due to the lack of data in active distribution networks, employing new accurate measurement devices like phasor measurement units (PMUs) and micro-PMUs with a high reporting rate becomes an inevitable choice for the future vision of distribution systems. As a result, different algorithms have been presented to optimally place PMUs cost-effectively based on the estimation errors of the distribution state estimation (DSE) results. However, any component failure in measurement devices or communication links between sending ends and monitoring system of the distribution management system can significantly affect the DSE results. In response, this study introduces the reliability of satisfying accuracy constraints (RSACs) as an important requirement in the DSE problem. This reliability index along with the estimation errors of the DSE results are then used as two performance indices for determining the optimal number and location (configuration) of PMUs. Finally, the performance of the proposed algorithm in comparison with traditional approaches in different topologies and operating conditions of two considered ADNs is evaluated. The results proved that employing the RSACs in the optimal PMU placement problems leads to a PMU configuration with the optimum RSACs and also the cost and latency of the communication system between all configurations which have the same measurement cost and satisfy accuracy constraints of DSE results.