Weighted Gain Combining

The core principle assigns weights directly proportional to each node's instantaneous SNR. Nodes experiencing favorable channel conditions (high SNR) receive larger weights, while nodes in deep fades (low SNR) are effectively de-emphasized. This maximizes the contribution of reliable sensors and minimizes the corrupting influence of noisy measurements on the global test statistic. The optimal weight vector under the Neyman-Pearson criterion is derived from the ratio of signal variance to noise variance at each receiver.

The fusion center computes a weighted sum of the energy estimates reported by all cooperating nodes. Mathematically, the global test statistic T = Σ wᵢ · yᵢ, where yᵢ is the energy measurement from node i and wᵢ is its assigned weight. This linear operation preserves the Gaussian nature of the aggregated statistic under both hypotheses, enabling closed-form analysis of detection performance. The linearity also ensures computational simplicity at the fusion center, making it suitable for real-time implementation.

Weighted gain combining directly addresses the hidden node problem where a single sensor may be shadowed from a primary transmitter. By aggregating weighted measurements from geographically distributed nodes, the fusion center effectively synthesizes a virtual antenna array with spatial diversity. Even if several nodes are in deep fades, the high-SNR nodes dominate the weighted sum, ensuring the global probability of detection remains robust against correlated shadowing and multipath nulls.

Unlike hard decision fusion where nodes transmit only binary decisions, weighted gain combining preserves the full dynamic range of the test statistic. This soft information retention provides a significant detection sensitivity gain, typically 2-3 dB in required SNR for equivalent performance. The tradeoff is increased reporting channel bandwidth, as nodes must transmit quantized analog values rather than single bits. Quantized soft combining with 3-4 bits per report captures most of this gain while constraining overhead.

The reliance on weighted contributions makes the fusion rule susceptible to Spectrum Sensing Data Falsification (SSDF) attacks. A malicious node reporting an artificially inflated energy value with a plausible SNR estimate can dominate the weighted sum and corrupt the global decision. Mitigation requires integrating reputation management mechanisms that dynamically adjust weights based on historical reporting consistency, effectively down-weighting nodes whose reports deviate persistently from the consensus.

Optimal weight design requires knowledge of both sensing channel and reporting channel statistics. While sensing channel SNR determines the reliability of the local measurement, imperfect reporting channels introduce errors between the node's computed value and the fusion center's received value. Robust weighted gain combining schemes incorporate reporting channel state information into the weight calculation, often using a minimum mean square error criterion to jointly optimize for both sensing fidelity and transmission reliability.