Skip to Main content Skip to Navigation

Gestion des interférences dans les systèmes MIMO massifs

Abstract : This thesis made it possible to work on the efficiency of a channel of massive MIMO systems for which it is necessary to determine the throughput at the Uplink of the terminals present in their respective cells. As an assumption, the frequency band in TDD mode is reused in each cell. All symbols are propagated asynchronously by the terminals present in the cells, not effectively preventing intra- and inter-symbol interactions at the base stations. These signals encounter many obstacles on their path that lead to delays, signal losses (destructive), signal regenerations (constructive) with various types of modulation (amplitude, frequency, phase), etc. The path loss in the channel is highlighted with the different values taken by the attenuation coefficient chosen during the simulations. Faced with this situation, it was necessary to look for the best and most robust channel estimator at a given consistency time. The MMSE (Minimum Mean Square Error) method is used, compared to others. For the performance of massive MIMO systems, we have focused on antenna diversity methods (N-order diversity), coding methods, OFDMA access methods and equalization methods to show that effectively using multiple antennas at base stations improves and contributes to the desired rate gains. With massive MIMO systems, we have shown that antennar contribution is well recognized in interference management. An algorithm for calculating the flow rate at the Uplink was developed using three conventional receivers: the MRC (Maximum Ratio Combiner), the ZF (Zero-Forcing) and the MMSE (Minimum Mean Square Error). The simulations made it possible to compare the different approaches. By varying the contamination power of the pilot symbols, we observe the convergence of the ZF and MMSE curves. If the number of L cells increases, we find that the higher the contamination power of the pilot symbols (pp), the lower the capacity in the channel. After several iterations, our algorithm converges to an asymptote (stationary and linear regime) where the samples at the detector output approach the transmitted data sequence. The SINR obtained with conventional detectors allows the calculation of the respective flows in the channel with the SHANNON theorem.
Complete list of metadatas

Cited literature [113 references]  Display  Hide  Download
Contributor : Abes Star :  Contact
Submitted on : Wednesday, March 13, 2019 - 5:46:26 PM
Last modification on : Thursday, April 25, 2019 - 1:21:47 AM
Long-term archiving on: : Friday, June 14, 2019 - 5:06:55 PM


Version validated by the jury (STAR)


  • HAL Id : tel-02066900, version 1



Bamba Sissokho. Gestion des interférences dans les systèmes MIMO massifs. Traitement du signal et de l'image [eess.SP]. Université de Limoges, 2019. Français. ⟨NNT : 2019LIMO0008⟩. ⟨tel-02066900⟩



Record views


Files downloads