Integrated sensing and communication (ISAC) technology endows users with environmental awareness capabilities, which will play a crucial role in future mobile edge computing (MEC) systems. In this paper, we consider the design of sensingassisted beam alignment and investigate resource management for a task-oriented mmWave integrated sensing, communication, and computing (ISCC) system, which can be formulated as a preference-agnostic multi-objective optimization problem. To solve this problem, we first introduce a multi-objective Markov decision process (MOMDP) to reformulate the original problem and innovatively propose a preference-agnostic multi-objective soft actor-critic (PA-MOSAC) algorithm. To demonstrate the effectiveness of our proposed system architecture and resource management algorithm, we also introduce a traditional mmWave MEC (T-MEC) system based on the same set of system parameters as a benchmark. The proximal policy optimization (PPO) algorithm, known for its robustness, is used to address resource management in the T-MEC system. Through a comprehensive comparative analysis of the two systems and algorithms, we discover that our proposed sensing-assisted beam alignment can reduce task execution delay by 25% with only 1% increase in energy consumption. We also verify the convergence of our proposed PA-MOSAC algorithm and demonstrate its superior performance over the benchmark scheme.

Covert communication enhances the security of information transmission by protecting communication behavior, which has been widely investigated in various wireless communication systems. In this paper, we develop a covert Internet of things (IoT) network, where energy-limited IoT nodes selectively transmit information or artificial noise based on channel state information. Specifically, if the channel gains of all legitimate links are less than a given threshold, the energy-limited node emits artificial noise, otherwise, it transmits covert information. Meanwhile, we adopt the strategy of channel inversion power control (CIPC) to hide the location information of the nodes. Then, we derive Willie's detection error probability (DEP), propose an optimization problem for covert throughput, and obtain the optimal threshold and transmission energy. The results indicate that the probability of covert transmission significantly affects Willie's DEP. Moreover, the average covert throughput can be maximized by a relatively small threshold.

In order to overcome the vicious competition between different high-volume services, we study the wireless resource sharing problem in the transmission process of the MEC-caching-coexist (MCCe) system with the capability of mmWave communications. The multiobjective Markov decision process (MOMDP) is introduced to model the task scheduling and resource allocation problem for the mmWave links, which aims to minimize the transmission delay and energy consumption simultaneously. Note that, for practical consideration, the exact channel information of all links are not known. We propose a novel multiobjective deep reinforcement learning with discrete-continuous hybrid action space (MODRL/HA) algorithm. In particular, the envelope updated design (EUD) is designed to realize the multiobjective optimization from the perspective of the Bellman operator. On the other hand, the parameterized network design (PND) is developed to deal with the hybrid action space of discrete task scheduling and continuous beamwidth and power variables. Our simulations show that, the MODRL/HA algorithm can improve 22% performance in terms of the tradeoff between delay and energy consumption compared with the benchmark schemes, which are original deep deterministic policy gradient (DDPG) and multiobjective DDPG (MODDPG) algorithms.

Upon exploiting massive spectrum resources, millimeter-wave (mmW) communication can significantly improve the transmission rate of mobile-edge computing (MEC) offloading, whereas the directional mmW links are constrained by shrunk beam coverage and demand extra phase for beam alignment. To enhance the accessing efficiency, we develop the nonorthogonal multiple access (NOMA) scheme-based mmW MEC mechanism, namely, NOMA-mmW MEC, therefore motivating to minimize the average delay of the MEC offloading, by jointly optimizing the beamwidth, user equipment (UE) scheduling, and transmit power. To tackle the mixed-integer nonlinear programming (MINLP) problem of delay minimization, we develop the alternative optimization (AO) approach-based RA scheme, namely, AO-RA, to obtain the close-optimum solutions. In the AO-RA scheme, we propose the matrix control many-to-one with externality (MC-M2OE) algorithm, to find the best UE scheduling for the NOMA groupings of different types of UEs. Upon the above, we further design the joint beamwidth and transmit power (JBTP) algorithm, which determines the optimal beamwidth and transmit power for the MEC offloading transmissions. Our simulation results show the effectiveness of the proposed AO-RA scheme in minimizing the offloading delay, where our MC-M2OE and JBTP algorithms can significantly outperform the existing approaches. From the simulation results, we may conclude that it needs to carefully address the tradeoff between beam alignment overhead and transmission gain while properly balancing the loading among different NOMA groups, for the practical consideration of NOMA-mmW MEC technology.

With exploiting massive spectrum resources, millimeter wave (mmWave) communications significantly improve the offloading capability for future mobile edge computing (MEC) techniques, which however is constrained by blockage problem in dynamic environments. In this paper, we study the resource allocation problem for the conceived mmWave MEC system with dynamic offloading process, in which the UEs are characterized by being mobile and having the imperfect knowledge of the offloading tasks coming. By introducing the multi-objective Markov decision process (MOMDP), the resource allocation problem is modeled by simultaneously minimizing the delay and energy consumption, where jointly considering the multi-beam assignment (mBA) and beamwidth and power optimization (BPO). To tackle this problem, we innovatively propose a matching-aided-learning (MaL) resource allocation scheme, with the aid of a learnable weight based attention mechanism (LW-AM) for adapting the dynamic offloading process. In particular, our MaL scheme includes many-to-one matching (M2O-M) based mBA algorithm and deep deterministic policy gradient (DDPG) based BPO algorithm, which are executed iteratively and converge with relatively low number of iterations. The simulation results show the practical value of the proposed MaL, which can approach the performance of benchmark scheme with perfect knowledge of offloading tasks.

Toward 6G networks, such as virtual reality (VR) applications, Industry 4.0, and automated driving, demand mobile-edge computing (MEC) techniques to offload computing tasks to nearby servers, which, however, causes fierce competition with traditional communication services. On the other hand, by introducing millimeter wave (mmWave) communication, it can significantly improve the offloading capability of MEC, enabling low latency and high throughput. For this sake, this article investigates the resource management for the offload transmission of the mmWave MEC system, when considering the data transmission demands from both communication-oriented users (CM-UEs) and computing-oriented users (CP-UEs). In particular, the joint consideration of user pairing, beamwidth allocation, and power allocation is formulated as a multiobjective problem (MOP), which includes minimizing the offloading delay of CP-UEs and maximizing the transmission rate of CM-UEs. By using the$\epsilon $-constraint approach, the MOP is converted into a single-objective optimization problem (SOP) without losing Pareto optimality, and then the three-stage iterative resource allocation algorithm is proposed. Our simulation results show that the gap between Pareto front generated by the three-stage iterative resource allocation algorithm and the real Pareto front is less than 0.16%. Furthermore, the proposed algorithm with much lower complexity can achieve the performance similar to the benchmark scheme of NSGA-II, while significantly outperforms the other traditional schemes.

This paper investigates the resource allocation in millimeter wave ultra dense networks, in which the beam assignment and sub-band allocation are jointly considered. Motivating to maximize the sum rate of the network conceived, the optimization problem is formulated as a mixed integer non-linear programing (MINLP) problem, which involves allocating the novel three-dimensional resource blocks (RBs) defined in beam (B), time (T), and frequency (F) dimension, respectively. To tackle the formulated MINLP problem, we propose the low-complexity resource allocation scheme, including the so-called best option first (BOF) beam assignment algorithm, and the many-to-one matching with externalities (M2O-ME) sub-band allocation algorithm. In particular, the BOF beam assignment algorithm is first carried out to coordinate the RBs in terms of T- and B-dimension. Then, with the aid of the mechanism of many-to-one with externalities, the M2O-ME sub-band algorithm is implemented to find the optimal sub-band allocation (i.e. RB allocation in F-dimension) solution. Finally, our simulation results show that the proposed resource allocation scheme can significantly outperform the existing schemes in terms of sum rate of the networks. Therefore, we can conclude that the proposed resource allocation scheme can be considered as a promising candidate for practical ultra dense small-cell networks with mmWave capability.