Curriculum Learning for Goal-Oriented Semantic Communications with a Common Language

Goal-oriented semantic communication will be a pillar of next-generation wireless networks. Despite significant recent efforts in this area, most prior works are focused on specific data types (e.g., image or audio), and they ignore the goal and effectiveness aspects of semantic transmissions. In contrast, in this paper, a holistic goal-oriented semantic communication framework is proposed to enable a speaker and a listener to cooperatively execute a set of sequential tasks in a dynamic environment. A common language based on a hierarchical belief set is proposed to enable semantic communications between speaker and listener. The speaker, acting as an observer of the environment, utilizes the beliefs to transmit an initial description of its observation (called event) to the listener. The listener is then able to infer on the transmitted description and complete it by adding related beliefs to the transmitted beliefs of the speaker. As such, the listener reconstructs the observed event based on the completed description, and it then takes appropriate action in the environment based on the reconstructed event. An optimization problem is defined to determine the perfect and abstract description of the events while minimizing the transmission and inference costs with constraints on the task execution time and belief efficiency. Then, a novel bottom-up curriculum learning (CL) framework based on reinforcement learning is proposed to solve the optimization problem and enable the speaker and listener to gradually identify the structure of the belief set and the perfect and abstract description of the events. Simulation results show that the proposed CL method outperforms traditional RL in terms of convergence time, task execution cost and time, reliability, and belief efficiency.

Spectrum Trading for Device-to-Device Communication In Cellular Networks using Incomplete Information Bandwidth-Auction Game

Device-to-device (D2D) communication that allows proximity users to communicate directly has been recently proposed to improve spectral efficiency of cellular networks. In this paper, we assume a cellular network consisting of multiple cellular user equipments (CUEs), which are the primary users, and a cognitive D2D pair, which is the secondary user. The D2D pair needs a bandwidth for data transmission that can be obtained via spectrum trading. We introduce a bandwidth-auction game for the spectrum trading problem. The base station (BS) and CUEs are able to sell their spectrum or share it with the D2D pair, which allows the D2D pair to operate in orthogonal sharing, cellular, or non-orthogonal sharing (NOS) modes. Operation of the D2D pair in the NOS mode causes interference to the CUEs, which is possible under low interference condition. In the auction, the D2D pair can buy its required spectrum from three different service providers (SPs) corresponding to each mode that operateon different frequency spectrums. The D2D pair bids a price bandwidth demand curve and the SPs offer a price-demand supply curve. Since each player is not aware of the strategy of other players in practical scenarios, the game is assumed to be an incomplete information repeated one. A best response based learning method is proposed for the decision making procedure of all players, the D2D pair and SPs. It is shown that the proposed method converges to the Nash equilibrium (NE) point of the game more rapidly than the state-of-the-art methods when the game is played repeatedly. The sensitivity of the proposed method to the learning rate variable is also less than the state-of-the-art methods and hence can be considered as a robust one.

Common Language for Goal-Oriented Semantic Communications: A Curriculum Learning Framework

Semantic communications will play a critical role in enabling goal-oriented services over next-generation wireless systems. However, most prior art in this domain is restricted to specific applications (e.g., text or image), and it does not enable goal-oriented communications in which the effectiveness of the transmitted information must be considered along with the semantics so as to execute a certain task. In this paper, a comprehensive semantic communications framework is proposed for enabling goal-oriented task execution. To capture the semantics between a speaker and a listener, a common language is defined using the concept of beliefs to enable the speaker to describe the environment observations to the listener. Then, an optimization problem is posed to choose the minimum set of beliefs that perfectly describes the observation while minimizing the task execution time and transmission cost. A novel top-down framework that combines curriculum learning (CL) and reinforcement learning (RL) is proposed to solve this problem. Simulation results show that the proposed CL method outperforms traditional RL in terms of convergence time, task execution time, and transmission cost during training.