Large language models (LLMs) face significant challenges in specialized domains like telecommunication (Tele-com) due to technical complexity, specialized terminology, and rapidly evolving knowledge. Traditional methods, such as scaling model parameters or retraining on domain-specific corpora, are computationally expensive and yield diminishing returns, while existing approaches like retrieval-augmented generation, mixture of experts, and fine-tuning struggle with accuracy, efficiency, and coordination. To address this issue, we propose Telecom mixture of models (TeleMoM), a consensus-driven ensemble framework that integrates multiple LLMs for enhanced decision-making in Telecom. TeleMoM employs a two-stage process: proponent models generate justified responses, and an adjudicator finalizes decisions, supported by a quality-checking mechanism. This approach leverages strengths of diverse models to improve accuracy, reduce biases, and handle domain-specific complexities effectively. Evaluation results demonstrate that TeleMoM achieves a 9.7% increase in answer accuracy, highlighting its effectiveness in Telecom applications.

Millimeter-wave (mmWave) technology is increasingly recognized as a pivotal technology of the sixth-generation communication networks due to the large amounts of available spectrum at high frequencies. However, the huge overhead associated with beam training imposes a significant challenge in mmWave communications, particularly in urban environments with high background noise. To reduce this high overhead, we propose a novel solution for robust continuous-time beam tracking with liquid neural network, which dynamically adjust the narrow mmWave beams to ensure real-time beam alignment with mobile users. Through extensive simulations, we validate the effectiveness of our proposed method and demonstrate its superiority over existing state-of-the-art deep-learning-based approaches. Specifically, our scheme achieves at most 46.9% higher normalized spectral efficiency than the baselines when the user is moving at 5 m/s, demonstrating the potential of liquid neural networks to enhance mmWave mobile communication performance.

Reconfigurable intelligent surfaces (RISs) have become a promising technology to meet the requirements of energy efficiency and scalability in future six-generation (6G) communications. However, a significant challenge in RISs-aided communications is the joint optimization of active and passive beamforming at base stations (BSs) and RISs respectively. Specif-ically, the main difficulty is attributed to the highly non-convex optimization space of beamforming matrices at both BSs and RISs, as well as the diversity and mobility of communication scenarios. To address this, we present a greenly gradient based meta learning beamforming (GMLB) approach. Unlike traditional deep learning based methods which take channel information directly as input, GMLB feeds the gradient of sum rate into neural networks. Coherently, we design a differential regulator to address the phase shift optimization of RISs. Moreover, we use the meta learning to iteratively optimize the beamforming matrices of BSs and RISs. These techniques make the proposed method to work well without requiring energy-consuming pretraining. Simulations show that GMLB could achieve higher sum rate than that of typical alternating optimization algorithms with the energy consumption by two orders of magnitude less.

Reconfigurable intelligent surface (RIS) has become a promising technology to realize the programmable wireless environment via steering the incident signal in fully customizable ways. However, a major challenge in RIS-aided communication systems is the simultaneous design of the precoding matrix at the base station (BS) and the phase shifting matrix of the RIS elements. This is mainly attributed to the highly non-convex optimization space of variables at both the BS and the RIS, and the diversity of communication environments. Generally, traditional optimization methods for this problem suffer from the high complexity, while existing deep learning based methods are lacking in robustness in various scenarios. To address these issues, we introduce a gradient-based manifold meta learning method (GMML), which works without pre-training and has strong robustness for RIS-aided communications. Specifically, the proposed method fuses meta learning and manifold learning to improve the overall spectral efficiency, and reduce the overhead of the high-dimensional signal process. Unlike traditional deep learning based methods which directly take channel state information as input, GMML feeds the gradients of the precoding matrix and phase shifting matrix into neural networks. Coherently, we design a differential regulator to constrain the phase shifting matrix of the RIS. Numerical results show that the proposed GMML can improve the spectral efficiency by up to 7.31%, and speed up the convergence by 23 times faster compared to traditional approaches. Moreover, they also demonstrate remarkable robustness and adaptability in dynamic settings.