In this paper, we present an innovative federated learning (FL) approach that utilizes Kolmogorov-Arnold Networks (KANs) for classification tasks. By utilizing the adaptive activation capabilities of KANs in a federated framework, we aim to improve classification capabilities while preserving privacy. The study evaluates the performance of federated KANs (F-KANs) compared to traditional federated Multi-Layer Perceptrons (F-MLPs) on classification task. The results show that the F-KANs model significantly outperforms the F-MLP model in terms of accuracy, precision, recall, F1 score and stability, and achieves better performance, paving the way for more efficient and privacy-preserving predictive analytics.

Low-Earth orbit (LEO) satellite networks, being a promising component of non-terrestrial networks (NTN), are meant to provide seamless and ubiquitous global connectivity. However, the high mobility of the LEO satellites results in frequent handovers (HOs), poses significant challenges to maintaining service continuity due to the interruption time experienced during the HO execution phase, which is particularly critical for high reliability communications (HRC). In this aspect, 3GPP proposed the dual active protocol stack (DAPS) HO mechanism for terrestrial networks (TN), a soft HO approach that enables the user equipment (UE) to maintain the simultaneous connection between both the source and target gNBs during the HO execution phase, potentially achieving a zero handover interruption time (HIT). However, DAPS HO is still not supported for NTN in latest 3GPP Release 18, likely due to rapid orbital movement of satellites. Moreover, DAPS HO requires synchronization between source and target nodes during the handover process, a condition that is challenging to identify at all times. Inspired by this challenge, this paper proposes a network-controlled algorithm that enhances the feasibility of DAPS HO in LEO satellites network, thereby minimizing the HIT over a given observation period. The proposed algorithm employs a sequential decision-making process within a predefined window to make the decision on configuring the DAPS HO. Simulation results demonstrate that the proposed approach significantly reduces the HIT, achieving a 80.28% reduction as compared to baseline HO scheme.

The integration of non-terrestrial networks into the 5G ecosystem is mainly driven by the possibility of provisioning service in remote areas. In this context, the advent of flying base stations at the low Earth orbit (LEO) will enable anywhere and anytime connectivity. To materialize this vision, it is of utmost importance to improve radio protocols with the aim of allowing direct satellite access. Bearing this aspect in mind, we present a new random access signal, which is based on the filter bank multicarrier (FBMC) waveform, and a computationally efficient detection scheme. The proposed solution outperforms the standardized access scheme based on single-carrier frequency division multiplexing (SC-FDM), by reducing out-of-band (OOB) emissions and reducing the missed detection probability in presence of very high carrier frequency offset (CFO), which is inherent to LEO satellite systems. The improvement is related to the fine frequency resolution of the detector and the use of pulse shaping techniques. Interestingly, the FBMC-based random access signal achieves a high level of commonality with 5G new radio, as the preamble generation method and the time-frequency allocation pattern can be kept unchanged. Concerning the practical implementation aspects, the complexity of the detector is similar in both SC-FDM and FBMC.

A novel coding strategy is proposed for a broadcast setting with two transmitter (TX) antennas and two single-antenna receivers (RX). The strategy consists of using space-time block coding to send a common message (to be decoded by both RXs) across the two TX antennas, while each TX antenna also sends a private message to one of the RXs. The relative weight of the private and common messages from each TX antenna is tuned to maximize the instantaneous achievable sum-rate of the channel. Closed-form expressions for the optimal weight factors are derived. In terms of the generalized degrees of freedom (GDoF) metric, the new scheme is able to achieve the sum-GDoF with finite precision channel state information at the transmitter (CSIT) of the two user broadcast channel. Moreover, as opposed to the existing rate-splitting schemes, the proposed scheme yields instantaneous achievable rates that are independent of the channel phases. This property is instrumental for link adaptation when only magnitude CSIT is available. Our numerical results indeed demonstrate the superiority of the scheme for the 2-user setting in case of magnitude CSIT. Extension to a more general K-user scenario is briefly discussed.

This paper investigates the application of multicast non-orthogonal multiple access (MC-NOMA) schemes to the forward link of a satellite communication system. In multicast transmission each frame contains information of multiple users. To benefit from the theory developed in NOMA, the proposed scheme creates two groups of users within each beam. The analysis conducted in this work reveals that the user grouping has an impact on the performance. In the light of this observation, power allocation and user clustering techniques have been derived to either maximize the sum-rate or achieve max-min fairness. The numerical simulation results show that MC-NOMA outperforms multicast orthogonal multiple access (MC-OMA) schemes, where different groups are served in orthogonal resources. Moreover, the gain of MC-NOMA over the MC-OMA becomes more prominent as number of users per group and the transmit power increases. The results show the minimum-rate and the sum-rate of MC-NOMA can be increased by a factor 2 and 1.45 with respect to MC-OMA, respectively.

Non-Coherent Rate-Splitting (NCRS) was recently proposed as a practical multiuser coding and decoding scheme to increase the spectral efficiency of multibeam satellite communication systems. In this paper, we further study the practical realization of NCRS. We propose a modified coding scheme (NCRS*) that is robust to a nonzero time offset among beams. In NCRS*, as opposed to NCRS, the beams send independently channel encoded and modulated waveforms. We assess the performance of NCRS* in terms of the achievable rate region. It is shown that NCRS* performs worse than NCRS, but better than or comparable to other competing schemes, which, as opposed to NCRS*, require flexible bandwidth allocation or perfect synchronization at the transmitter. We also propose a new N-MAP algorithm for the practical implementation of NCRS* receivers. Similar to the existing UMAP algorithm, N-MAP takes into account the modulation used by, and the time offset between, the signals received from the different beams. In most cases, however, N-MAP has a significantly lower complexity than U-MAP.

Next generation satellite payload technology is expected to provide digital processing capabilities. This will pave the way to regard satellites as flying base stations. However, the initial access procedure must be improved. In this work we focus on non-geostationary earth orbit (NGEO) satellite networks. In this scenario, the random access preamble signal and the detection must be robust to large carrier frequency offsets (CFOs). Towards this end, we investigate the adoption of the pruned discrete Fourier transform spread filter bank multicarrier waveform. The proposed design is suitable for the access scheme of forthcoming 5G-based NGEO satellite communications. The reason is twofold. First, it improves the spectral confinement with respect to the standard single-carrier frequency-division multiplexing (SC-FDM) waveform. Second, it achieves a high level of commonality with 5G new radio, by keeping unchanged the subcarrier spacing, the slot duration and the preamble sequence. Remarkably, the new design allows the straightforward application of non-coherent post detection integration (NCPDI) techniques, which divide the correlation in blocks. Numerical results show that the proposed solution reduces out-of-band emissions and the missed detection probability in presence of CFO, with respect to the conventional approach based on SC-FDM and preamble detection with full-length correlation.

This paper considers interference mitigation techniques in the forward link of multibeam satellite systems. In contrast to previous works, either devoted to receiver interference mitigation (e.g. multiuser detection) or transmitter interference mitigation (precoding), this work evaluates the achievable rates of the joint combination of both techniques. On the one hand, precoding cannot properly mitigate all the inter- beam interference while maintaining a sufficiently high signal-to-noise ratio. On the other hand, the receiver cost and complexity exponentially increases with the number of signals to be simultaneously detected. This highlights that the receiver cannot deal with all the interferences so that in general only 2 signals are jointly detected. As a result, the use of precoding within a coverage area jointly with multiuser detection can both benefit from each other and extremely increase the achievable rates of the system. This is numerically evaluated in a close-to-real coverage area considering simultaneous non-unique decoding strategies. The results show the benefits of this joint scheme that eventually can increase the current precoding performance a 23%.

This paper evaluates the capacity of weighted circularly convolved FBMC/OQAM systems. A rigorous mathematical model is derived to calculate the increase of capacity that can be obtained thanks to the lattice structure of the modulation and the exploitation of the intrinsic interference. The numerical results reveal that a signal to noise ratio gain of 2dB is obtained in one resource block of 12 subcarriers, translating into a capacity increase of 11% with respect to OFDM, which nuances some previous results on this topic in the literature.

This paper investigates the application of filter bank multicarrier modulation based on the OQAM (FBMC/OQAM) to multiple-input-multiple-output (MIMO) systems. Existing solutions guarantee satisfactory performance when the streams multiplexed on each subcarrier (S) and the number of transmit (NT) and receive (NR) antennas are related as S = min (NT, NR). When ST, NR), the techniques presented in previous works either exhibit an error floor or perform much worse than orthogonal frequency division multiplexing (OFDM). To make progress towards the combination of FBMC/OQAM with MIMO we propose a two-step approach and a coordinated beamforming algorithm to design the transmit and the receive processing. Numerical results show that the two-step method provides similar bit error rate (BER) as OFDM when S + 1 = NT= NR. Resorting to the coordinated beamforming solution, which is based on an iterative method, the application of FBMC/OQAM is extended to the general case ST, NR). Hence, the techniques presented in this paper demonstrate that FBMC/OQAM can achieve practically the same BER as OFDM with an increased spectral efficiency and a significantly decreased out-of-band radiation, which is an important advantage for non-contiguous spectrum allocations.

This paper addresses the design of MIMO precoding and decoding techniques for the filter bank multicarrier (FBMC) modulation. Existing solutions give satisfactory performance in scenarios with high coherence bandwidth channels. With the aim of increasing the robustness against the channel frequency selectivity, we have rethought the problem, which results in a new subband processing. Simulation-based results show that the proposed solution can achieve similar bit error rates as the OFDMsolution, while the spectral efficiency is increased. These results are theoretically justified. As a conclusion, FBMC becomes attractive in frequency selective channels not only because it relaxes the frame synchronization with respect to OFDM, but also because it presents spatial multiplexing competitive results.

Several studies have revealed that OFDM is outperformed in several areas by the filter bank based multicarrier (FBMC) modulation thanks to pulse shaping techniques. Nevertheless, the combination of FBMC with multi-antenna architectures is nontrivial. Aiming at making progress towards this direction we study MIMO precoding techniques for the MISO broadcast channel (MISO-BC). Due to its superior performance over linear precoding techniques we have focused on the spatial Tomlinson Harashima precoder (STHP). The analysis carried out in this paper concludes that the conventional STHP performs poorly when it is combined with FBMC systems since it is not able to cope with ISI and ICI. To combat the interferences we propose a novel subband processing based on the STHP concept. The numerical results show that FBMC and OFDM almost give the same performance in terms of BER. However, FBMC is able to achieve spectral efficiency values that are 31% higher in comparison to the ones obtained in OFDM when 20% of the symbol duration corresponds to the cyclic prefix.

In this letter the discrete rate maximization problem is investigated for FBMC/OQAM. The analysis reveals that if there is crosstalk certain bit allocations violate the power constraints. Aiming at ensuring the feasibility along with alleviating the complexity we have devised a novel iterative algorithm, which always converges. Simulation-based results show that the proposed algorithm performs close to the upper bound for high-coherence bandwidth channels. Under highly frequency selective channels the existing algorithms are not able to guarantee the target SER whereas our approach guarantees the QoS.

Build upon the OFDM/OQAM scheme this paper addresses the design of linear transmit and receive filters. Aiming at enhancing the robustness against multipath fading, two novel techniques are proposed. The first one jointly designs transmit and receive filters under the MMSE criterion. The second scheme follows a ZF approach. In this case transmit and receive filters are independently devised. Pre-equalizers obtained by both strategies are subjected to power constraints. Simulation-based results show that the two proposed techniques benefit from having perfect channel knowledge at both ends of the link to outperform the Alamouti scheme. The best performance in terms of BER is given when transmit and receive filters are jointly designed.