Pilot-Power Control for Interference Suppression in Reciprocity-Based Systems: Separate Pilot Design for Channel Estimation and Interference Suppression
2026 (English)Independent thesis Advanced level (degree of Master (Two Years)), 28 HE credits
Student thesisAlternative title
Pilotstyrning för interferensundertryckning i reciprocitetsbaserade system : Separat pilotdesign för kanalestimering och interferensundertryckning (Swedish)
Abstract [en]
This thesis investigates pilot-power control in multi-cell massive multiple-input multiple-output (MIMO) systems, withparticular focus on pilot-contaminated settings where pilot signaling affects not only channelestimation but also spatial interference suppression. The work considers both conventionalpilot-based precoding and residual-based precoding with separate pilot signals for channelestimation and interference suppression.
The analysis is carried out under an uncorrelated Rayleigh fading model using closed-formsignal-to-interference-plus-noise ratio (SINR) expressions, deterministic-equivalent approximations, and Monte Carlo simulations.Pilot powers are optimized numerically with respect to either sum spectral efficiency(SE) ormax-min fairness, and the resulting allocations are compared with full-power transmissionand with heuristic pilot-power rules based only on large-scale fading.
The results show that full pilot power is optimal when pilot contamination is absent. Whenpilot reuse is present, pilot-power optimization becomes beneficial and reveals a cleartrade-off between throughput and fairness. In the residual-based setting, the optimizedchannel-estimation and interference-suppression pilot powers exhibit different behaviors,which indicates that the two pilot signals should generally not be controlled by the samerule. The results also show that a low-complexity large-scale-fading-based heuristic can outperform thefull-power baseline while remaining close to the optimized throughput-orientedsolution.At the same time, the residual-based optimization relies on approximate surrogate models.The corresponding results should therefore beinterpreted primarily as qualitative rather than exact. Overall, the thesis shows thatpilot-power control in interference-aware massive MIMO is not only a channel-estimationproblem, but also an interference-management problem, and that large-scale-fading-basedheuristics constitute a promising practical alternative to repeated numerical optimization.
Place, publisher, year, edition, pages
2026. , p. 92
Keywords [en]
Massive MIMO, pilot-power control, pilot contamination, inter-cell interference, interference suppression, spectral efficiency
National Category
Electrical Engineering, Electronic Engineering, Information Engineering
Identifiers
URN: urn:nbn:se:liu:diva-227060ISRN: LiTH-ISY-EX--26/5828--SEOAI: oai:DiVA.org:liu-227060DiVA, id: diva2:2094366
External cooperation
Ericsson
Subject / course
Applied Mathematics
Presentation
2026-06-01, ISY Systemet, Mäster Mattias väg B-huset, Linköping, 13:15 (English)
Supervisors
Examiners
2026-08-242026-08-212026-08-24Bibliographically approved