Sphere Decoding with BER Enhancement in MIMO Systems: A Radius Prediction Approach with Wiener Filtering

Sphere decoding is a well-established technique for signal detection in MIMO systems, with its effectiveness heavily reliant on precise sphere radius prediction. To balance computational complexity and detection accuracy, existing methods such as Singular Value Decomposition, Chi-Square distribution, and Condition Number analysis are employed to forecast the sphere radius. However, accurate prediction alone may not fully mitigate noise and interference in the transmitted signal. To overcome these difficulties, Wiener filtering is employed to the transmitted signal as a pre-processing step before sphere decoding to further enhance performance. This filtering effectively reduces noise and interference, leading to significant improvements in BER. The proposed method was validated through simulations using a 10x10 MIMO configuration with Binary Phase Shift Keying modulation. The results demonstrated notable BER enhancements compared to traditional sphere decoding techniques without Wiener filtering. The results highlight how the integration of Wiener filtering with sphere decoding enables reliable and efficient signal detection in MIMO systems.