The most recent technology of Wireless Sensor Networks (WSNs) has infiltrated into real time use such as in environmental monitoring, smart cities, industrial automation and disaster management. They however are very vulnerable to network congestion due to limited bandwidth, burstiness of data traffic, and unbalanced sensor node load. The effects of congestion are the use of more energy, higher latency, loss of data, and the reoccurrence of packets; hence, less reliability of the network and a reduced lifetime. The traditional congestion control schemes (rate-limiting schemes or fixed set of adjustments on transmission power) fail to consider the dynamic traffic situation. They either overload the network beyond normal capacity or cause instability in the quality of links by cutting the power to unsustainable levels. Consequently, existing solutions do not assure the presence of the best trade-offs between energy consumption, throughput, and reliability when using WSNs with limited resources. In order to address the restricted capabilities, this paper introduces the concept of the Dynamic Transmission Power and Data Rate Balancing Strategy (DTP-DRB) that dynamically decreases the congestion effects by balancing the data rate and the transmission power together depending on the actual network situation. The proposed method uses Hybrid Congestion Index (HCI), calculated using the measures of the buffer congestion, packet delay and link quality to determine the degree of congestion. This algorithm is based on a two-fold optimisation which seeks to integrate a lightweight reinforcing learning where the algorithm dynamically decides on reducing the data rate, altering the transmission power, or deviating a traffic path. This leads to low energy wastages, network balance and efficient data delivery. The DTP-DRB is much higher than the traditional congestion schemes such as CODA and PCCP according to large-scale simulations. It provides 17% higher rate of sending packets, 28% higher network lifetime and enhancement of throughput by 20%, which reduces unnecessary retransmissions and hits. The findings suggest the potential of an adaptive power-controlled rates estimator, an effective and scalable congestion-sensitive framework of WSNs in receptive traffic and energy-restrained operating schemes.