Narrow beams in millimeter-wave (mmWave) communication introduce significant beam misalignment challenges. In this paper, we introduce MAMBA-X, an enhanced version of the MAMBA beam tracking scheme. Basically, MAMBA uses a restless multi-armed bandit framework to capture the dynamics of mmWave links by discounting the relevance of past observations using a “forgetting factor” ($\gamma_{1}$) and increases the weight of recent observations via a “boost factor” ($\gamma_{2}$). Because the original MAMBA uses fixed values for$\gamma_{1}$and$\gamma_{2}$, it cannot quickly adapt to variations in user mobility. Moreover, if the time between consecutive beam selection instances is large compared to channel dynamics, past observations become obsolete. To tackle these issues, we first use the concept of beam coherence time to establish a bound on the beam selection intervals. Secondly, we show that the performance of MAMBA depends primarily on the value of$\gamma_{1}$which, in turn, depends on UE mobility. We develop a Long Short-Term Memory (LSTM) model to dynamically predict and update the optimal value of$\gamma_{1}$. Through extensive simulations at 28 GHz and using publicly available 5G NR experimental dataset, we evaluate MAMBA-X. Our results indicate that the total delivered traffic is improved by up to 46.8% relative to the original MAMBA and 142% compared to the default beam management scheme in 5G NR.

Spectrum access/sharing algorithms for dynamic spectrum access (DSA) networks are often designed without accounting for adjacent-channel interference. In practice, guard bands are needed to prevent such interference. Introducing guard bands naturally constrains the effective use of the spectrum. In this work, we investigate the problem of assigning channels/powers to opportunistic transmissions, while accounting for such a constraint. Specifically, we propose a novel guard-band-aware channel assignment scheme for DSA systems. Our scheme reduces the number of required guard channels for a given transmission by exploiting the benefit of utilizing adjacent channels and considering already reserved guard channels. We analytically formulate the channel access problem as a joint power control and channel assignment optimization problem, with the objective of minimizing the required spectrum resource for a given CR transmission. We show that the optimization problem is a binary linear program (BLP), which is, in general, NP-hard. Accordingly, we present a near-optimal solution based on sequential fixing, where the binary variables are determined iteratively by solving a sequence of linear programs. Based on the proposed channel assignment algorithm, we develop an operational MAC protocol that enables DSA users to dynamically utilize the spectrum. The proposed protocol realizes our channel assignment algorithm in a distributed manner while relying only on information provided by the two communicating users. Simulation results are provided, which verify the effectiveness of our protocol and demonstrate the significant gain achieved through guard-band-aware channel assignment.

Spectrum sharing algorithms for cognitive radio networks (CRNs) are often designed ignoring adjacent-channel interference (i.e., interference from other transmissions operating on adjacent channels). In practice, such an assumption is unrealistic as guard bands are needed to prevent such interference. Introducing guard bands naturally constrains the effective use of the spectrum. In this work, we investigate the problem of assigning channels/powers to CR transmissions, while accounting for such a constraint. Specifically, we propose a novel guard-band-aware channel assignment scheme for CRNs. Our scheme reduces the number of required guard channels for a given transmission by exploiting the benefits of utilizing adjacent channels while considering already reserved guard channels. We analytically formulate the channel access problem as a joint power control and channel assignment optimization problem, with the objective of minimizing the required spectrum resource for a CR transmission. We show that the optimization problem is a binary linear program (BLP), which is, in general, NP-hard. Accordingly, we present a near-optimal solution based on a sequential fixing procedure. Simulation results are provided, which verify the accuracy of our algorithm and demonstrate the significant gain achieved through guard-band-aware channel assignment.