The authors propose a QoS based multi-objective optimization algorithm aiming at ensuring certain QoS levels in Wireless Mesh Networks . Some of the QoS parameters optimized are bandwidth, packet loss rates, delay and power consumption. Our approach targets WMSN QoS requirements instead. The ETX metric is used since it allows finding high throughput paths taking into account link loss ratios, links’ asymmetries, and interference among the successive links of a path. The authors in [14] have shown that low ETXs paths are also energy efficient. The selection of a link with a certain bandwidth does not guarantee that the path has a good throughput. Another difference was on the problem formulation presented. The authors of [17] presented a linear programming formulation, not making clear how the presented formulation is used by the multi-objective optimization algorithm. We modeled the routing problem as a multi-constrained QoS routing problem and consequently used multi-objective optimization algorithms to solve it. In addition, in [17], it was not clear how the MOEA algorithm was implemented. In contrast, we present and explain in detail how our MOEA algorithm was implemented, namely: (1) the population initialization process, (2) how genetic operators were used.