arXiv subjects

Zhifeng Zhao

Publications and source records attributed to Zhifeng Zhao.

At least 73 records · Page 4Linked to original sources

arXiv2022

AoI-based Temporal Attention Graph Neural Network for Popularity Prediction and Content Caching

Along with the fast development of network technology and the rapid growth of network equipment, the data throughput is sharply increasing. To handle the problem of backhaul bottleneck in cellular network and satisfy people's requirements about latency, the network architecture like information-centric network (ICN) intends to proactively keep limited popular content at the edge of network based on predicted results. Meanwhile, the interactions between the content (e.g., deep neural network models, Wikipedia-alike knowledge base) and users could be regarded as a dynamic bipartite graph. In this paper, to maximize the cache hit rate, we leverage an effective dynamic graph neural network (DGNN) to jointly learn the structural and temporal patterns embedded in the bipartite graph. Furthermore, in order to have deeper insights into the dynamics within the evolving graph, we propose an age of information (AoI) based attention mechanism to extract valuable historical information while avoiding the problem of message staleness. Combining this aforementioned prediction model, we also develop a cache selection algorithm to make caching decisions in accordance with the prediction results. Extensive results demonstrate that our model can obtain a higher prediction accuracy than other state-of-the-art schemes in two real-world datasets. The results of hit rate further verify the superiority of the caching policy based on our proposed model over other traditional ways.

cs.LG↗

arXiv2022

Adaptive Bit Rate Control in Semantic Communication with Incremental Knowledge-based HARQ

Semantic communication has witnessed a great progress with the development of natural language processing (NLP) and deep learning (DL). Although existing semantic communication technologies can effectively reduce errors in semantic interpretation, most of these solutions adopt a fixed bit length structure, along with a rigid transmission scheme, which is inefficient and lacks scalability when faced with different meanings and signal-to-noise ratio (SNR) conditions. In this paper, we explore the impact of adaptive bit lengths on semantic coding (SC) under various channel conditions. First, we propose progressive semantic hybrid automatic repeat request (HARQ) schemes that utilize incremental knowledge (IK) to simultaneously reduce the communication cost and semantic error. On top of this, we design a novel semantic encoding solution with multi-bit length selection. In this fashion, the transmitter employs a policy network to decide the appropriate coding rate, so as to secure the correct information delivery at the cost of minimal bits. Moreover, a specific denoiser is further introduced to reduce the semantic errors encountered in the transmission process according to the semantic characteristics of context. Extensive simulation results have been conducted to verify the effectiveness of the proposed solution.

eess.SY↗

arXiv2022

Rethinking Modern Communication from Semantic Coding to Semantic Communication

Modern communications are usually designed to pursue a higher bit-level precision and fewer bits while transmitting a message. This article rethinks these two major features and introduces the concept and advantage of semantics that characterizes a new kind of semantics-aware communication framework, incorporating both the semantic encoding and the semantic communication problem. After analyzing the underlying defects of existing semantics-aware techniques, we establish a confidence-based distillation mechanism for the joint semantics-noise coding (JSNC) problem and a reinforcement learning (RL)-powered semantic communication paradigm that endows a system the ability to convey the semantics instead of pursuing the bit level accuracy. On top of these technical contributions, this work provides a new insight to understand how the semantics are processed and represented in a semantics-aware coding and communication system, and verifies the significant benefits of doing so. Targeted on the next generation's semantics-aware communication, some critical concerns and open challenges such as the information overhead, semantic security and implementation cost are also discussed and envisioned.

eess.SP↗

arXiv2022

Reinforcement Learning-powered Semantic Communication via Semantic Similarity

We introduce a new semantic communication mechanism - SemanticRL, whose key idea is to preserve the semantic information instead of strictly securing the bit-level precision. Unlike previous methods that mainly concentrate on the network or structure design, we revisit the learning process and point out the semantic blindness of commonly used objective functions. To address this semantic gap, we introduce a schematic shift that learns from semantic similarity, instead of relying on conventional paired bit-level supervisions like cross entropy and bit error rate. However, developing such a semantic communication system is indeed a nontrivial task considering the non-differentiability of most semantic metrics as well as the instability from noisy channels. To further resolve these issues, we put forward a self-critic reinforcement learning (RL) solution which allows an efficient and stable learning on any user-defined semantic measurement, and take a step further to simultaneously tackle the non-differentiable semantic channel optimization problem via self-critic stochastic iterative updating (SCSIU) training on the decoupled semantic transceiver. We have firstly tested the proposed method in the challenging European-parliament dataset, which confirms the superiority of our method in revealing the semantic meanings, and better handling the semantic noise. Apart from the experimental results, we further provide an in-depth look at how the semantic model behaves, along with its superb generalization ability in real-life examples. An RL-based image transmission extension is also exemplified, so as to prove the generalization ability and motivate future discussion.

cs.LG↗

arXiv2022

Communication-Efficient Consensus Mechanism for Federated Reinforcement Learning

The paper considers independent reinforcement learning (IRL) for multi-agent decision-making process in the paradigm of federated learning (FL). We show that FL can clearly improve the policy performance of IRL in terms of training efficiency and stability. However, since the policy parameters are trained locally and aggregated iteratively through a central server in FL, frequent information exchange incurs a large amount of communication overheads. To reach a good balance between improving the model's convergence performance and reducing the required communication and computation overheads, this paper proposes a system utility function and develops a consensus-based optimization scheme on top of the periodic averaging method, which introduces the consensus algorithm into FL for the exchange of a model's local gradients. This paper also provides novel convergence guarantees for the developed method, and demonstrates its superior effectiveness and efficiency in improving the system utility value through theoretical analyses and numerical simulation results.

cs.LG↗

arXiv2021

Semantic Communication with Adaptive Universal Transformer

With the development of deep learning (DL), natural language processing (NLP) makes it possible for us to analyze and understand a large amount of language texts. Accordingly, we can achieve a semantic communication in terms of joint semantic source and channel coding over a noisy channel with the help of NLP. However, the existing method to realize this goal is to use a fixed transformer of NLP while ignoring the difference of semantic information contained in each sentence. To solve this problem, we propose a new semantic communication system based on Universal Transformer. Compared with the traditional transformer, an adaptive circulation mechanism is introduced in the Universal Transformer. Through the introduction of the circulation mechanism, the new semantic communication system can be more flexible to transmit sentences with different semantic information, and achieve better end-to-end performance under various channel conditions.

cs.CL↗

arXiv2021

Graph Attention Network-based Multi-agent Reinforcement Learning for Slicing Resource Management in Dense Cellular Network

Network slicing (NS) management devotes to providing various services to meet distinct requirements over the same physical communication infrastructure and allocating resources on demands. Considering a dense cellular network scenario that contains several NS over multiple base stations (BSs), it remains challenging to design a proper real-time inter-slice resource management strategy, so as to cope with frequent BS handover and satisfy the fluctuations of distinct service requirements. In this paper, we propose to formulate this challenge as a multi-agent reinforcement learning (MARL) problem in which each BS represents an agent. Then, we leverage graph attention network (GAT) to strengthen the temporal and spatial cooperation between agents. Furthermore, we incorporate GAT into deep reinforcement learning (DRL) and correspondingly design an intelligent real-time inter-slice resource management strategy. More specially, we testify the universal effectiveness of GAT for advancing DRL in the multi-agent system, by applying GAT on the top of both the value-based method deep Q-network (DQN) and a combination of policy-based and value-based method advantage actor-critic (A2C). Finally, we verify the superiority of the GAT-based MARL algorithms through extensive simulations.

cs.MA↗

arXiv2021

Stigmergic Independent Reinforcement Learning for Multi-Agent Collaboration

With the rapid evolution of wireless mobile devices, there emerges an increased need to design effective collaboration mechanisms between intelligent agents, so as to gradually approach the final collective objective through continuously learning from the environment based on their individual observations. In this regard, independent reinforcement learning (IRL) is often deployed in multi-agent collaboration to alleviate the problem of a non-stationary learning environment. However, behavioral strategies of intelligent agents in IRL can only be formulated upon their local individual observations of the global environment, and appropriate communication mechanisms must be introduced to reduce their behavioral localities. In this paper, we address the problem of communication between intelligent agents in IRL by jointly adopting mechanisms with two different scales. For the large scale, we introduce the stigmergy mechanism as an indirect communication bridge between independent learning agents, and carefully design a mathematical method to indicate the impact of digital pheromone. For the small scale, we propose a conflict-avoidance mechanism between adjacent agents by implementing an additionally embedded neural network to provide more opportunities for participants with higher action priorities. In addition, we present a federal training method to effectively optimize the neural network of each agent in a decentralized manner. Finally, we establish a simulation scenario in which a number of mobile agents in a certain area move automatically to form a specified target shape. Extensive simulations demonstrate the effectiveness of our proposed method.

cs.AI↗

arXiv2020

Learning to Prune in Training via Dynamic Channel Propagation

In this paper, we propose a novel network training mechanism called "dynamic channel propagation" to prune the neural networks during the training period. In particular, we pick up a specific group of channels in each convolutional layer to participate in the forward propagation in training time according to the significance level of channel, which is defined as channel utility. The utility values with respect to all selected channels are updated simultaneously with the error back-propagation process and will adaptively change. Furthermore, when the training ends, channels with high utility values are retained whereas those with low utility values are discarded. Hence, our proposed scheme trains and prunes neural networks simultaneously. We empirically evaluate our novel training scheme on various representative benchmark datasets and advanced convolutional neural network (CNN) architectures, including VGGNet and ResNet. The experiment results verify the superior performance and robust effectiveness of our approach.

cs.CV↗

arXiv2020

Internet of Intelligence: The Collective Advantage for Advancing Communications and Intelligence

The fifth-generation cellular networks (5G) has boosted the unprecedented convergence between the information world and physical world. On the other hand, empowered with the enormous amount of data and information, artificial intelligence (AI) has been universally applied and pervasive AI is believed to be an integral part of the six-generation cellular networks (6G). Consequently, benefiting from the advancement in communication technology and AI, we boldly argue that the conditions for collective intelligence (CI) will be mature in the 6G era and CI will emerge among the widely connected beings and things. Afterwards, we highlight the potential huge impact of CI on both communications and intelligence. In particular, we introduce a regular language (i.e., the information economy metalanguage) supporting the future collective communications to augment human intelligence and explain its potential applications in naming Internet information and pushing information centric networks forward. Meanwhile, we propose a stigmergy-based federated collective intelligence and demonstrate its achievement in a simulated scenario where the agents collectively work together to form a pattern through simple indirect communications. In a word, CI could advance both communications and intelligence.

cs.AI↗

arXiv2020

Intelligent Slicing of Radio Resource Control Layer for Cellular IoT: Design and Implementation

The cellular internet of things (CIoT) has become an important branch to cater various applications of IoT devices. Within CIoT, the radio resource control (RRC) layer is responsible for fundamental functionalities such as connection control and bearer establishment in radio access network (RAN). The emergence of various IoT scenarios and diversified service requirements have made both RAN slicing and intelligent control imperative requirement in RRC layer. This paper focuses on enhancing standardized capabilities of CIoT RRC layer, by designing and implementing a new architecture which accommodate RRC slicing and intelligent controller. The architecture aims to realize functionalities of creating, modifying, and deleting slices in RRC layer, while the intelligent controller is added to satisfy various and dynamic service requirements of different IoT devices smartly. The proposed architecture is further implemented on an open-source software platform OpenAirInterface (OAI), on top of which the effectiveness of RRC slicing is validated and one proof-of-concept case to adopt reinforcement learning to dynamically tune discontinuous reception parameters therein is presented. Simulation results have demonstrated the effectiveness of the proposed intelligent RRC slicing architecture.

cs.IT↗

arXiv2020

On the Capacity of Fractal D2D Social Networks with Hierarchical Communications

The maximum capacity of fractal D2D (device-to-device) social networks with both direct and hierarchical communications is studied in this paper. Specifically, the fractal networks are characterized by the direct social connection and the self-similarity. Firstly, for a fractal D2D social network with direct social communications, it is proved that the maximum capacity is $ Θ\left(\frac{1}{\sqrt{n\log n}}\right) $ if a user communicates with one of his/her direct contacts randomly, where $ n $ denotes the total number of users in the network, and it can reach up to $ Θ\left(\frac{1}{\log n}\right) $ if any pair of social contacts with distance $ d $ communicate according to the probability in proportion to $ d^{-β} $. Secondly, since users might get in touch with others without direct social connections through the inter-connected multiple users, the fractal D2D social network with these hierarchical communications is studied as well, and the related capacity is further derived. Our results show that this capacity is mainly affected by the correlation exponent $ε$ of the fractal structure. The capacity is reduced in proportional to $ \frac{1}{\log n} $ if $ 2<ε<3 $, while the reduction coefficient is $ \frac{1}{n} $ if $ ε>3 $.

cs.IT↗

arXiv2019

GAN-powered Deep Distributional Reinforcement Learning for Resource Management in Network Slicing

Network slicing is a key technology in 5G communications system. Its purpose is to dynamically and efficiently allocate resources for diversified services with distinct requirements over a common underlying physical infrastructure. Therein, demand-aware resource allocation is of significant importance to network slicing. In this paper, we consider a scenario that contains several slices in a radio access network with base stations that share the same physical resources (e.g., bandwidth or slots). We leverage deep reinforcement learning (DRL) to solve this problem by considering the varying service demands as the environment state and the allocated resources as the environment action. In order to reduce the effects of the annoying randomness and noise embedded in the received service level agreement (SLA) satisfaction ratio (SSR) and spectrum efficiency (SE), we primarily propose generative adversarial network-powered deep distributional Q network (GAN-DDQN) to learn the action-value distribution driven by minimizing the discrepancy between the estimated action-value distribution and the target action-value distribution. We put forward a reward-clipping mechanism to stabilize GAN-DDQN training against the effects of widely-spanning utility values. Moreover, we further develop Dueling GAN-DDQN, which uses a specially designed dueling generator, to learn the action-value distribution by estimating the state-value distribution and the action advantage function. Finally, we verify the performance of the proposed GAN-DDQN and Dueling GAN-DDQN algorithms through extensive simulations.

cs.LG↗

arXiv2019

Resource Awareness in Unmanned Aerial Vehicle-Assisted Mobile-Edge Computing Systems

This paper investigates an unmanned aerial vehicle (UAV)-assisted mobile-edge computing (MEC) system, in which the UAV provides complementary computation resource to the terrestrial MEC system. The UAV processes the received computation tasks from the mobile users (MUs) by creating the corresponding virtual machines. Due to finite shared I/O resource of the UAV in the MEC system, each MU competes to schedule local as well as remote task computations across the decision epochs, aiming to maximize the expected long-term computation performance. The non-cooperative interactions among the MUs are modeled as a stochastic game, in which the decision makings of a MU depend on the global state statistics and the task scheduling policies of all MUs are coupled. To approximate the Nash equilibrium solutions, we propose a proactive scheme based on the long short-term memory and deep reinforcement learning (DRL) techniques. A digital twin of the MEC system is established to train the proactive DRL scheme offline. Using the proposed scheme, each MU makes task scheduling decisions only with its own information. Numerical experiments show a significant performance gain from the scheme in terms of average utility per MU across the decision epochs.

cs.NI↗

arXiv2019

Secrecy Preserving in Stochastic Resource Orchestration for Multi-Tenancy Network Slicing

Network slicing is a proposing technology to support diverse services from mobile users (MUs) over a common physical network infrastructure. In this paper, we consider radio access network (RAN)-only slicing, where the physical RAN is tailored to accommodate both computation and communication functionalities. Multiple service providers (SPs, i.e., multiple tenants) compete with each other to bid for a limited number of channels across the scheduling slots, aiming to provide their subscribed MUs the opportunities to access the RAN slices. An eavesdropper overhears data transmissions from the MUs. We model the interactions among the non-cooperative SPs as a stochastic game, in which the objective of a SP is to optimize its own expected long-term payoff performance. To approximate the Nash equilibrium solutions, we first construct an abstract stochastic game using the channel auction outcomes. Then we linearly decompose the per-SP Markov decision process to simplify the decision-makings and derive a deep reinforcement learning based scheme to approach the optimal abstract control policies. TensorFlow-based experiments verify that the proposed scheme outperforms the three baselines and yields the best performance in average utility per MU per scheduling slot.

cs.NI↗

arXiv2019

Deep Reinforcement Learning with Discrete Normalized Advantage Functions for Resource Management in Network Slicing

Network slicing promises to provision diversified services with distinct requirements in one infrastructure. Deep reinforcement learning (e.g., deep $\mathcal{Q}$-learning, DQL) is assumed to be an appropriate algorithm to solve the demand-aware inter-slice resource management issue in network slicing by regarding the varying demands and the allocated bandwidth as the environment state and the action, respectively. However, allocating bandwidth in a finer resolution usually implies larger action space, and unfortunately DQL fails to quickly converge in this case. In this paper, we introduce discrete normalized advantage functions (DNAF) into DQL, by separating the $\mathcal{Q}$-value function as a state-value function term and an advantage term and exploiting a deterministic policy gradient descent (DPGD) algorithm to avoid the unnecessary calculation of $\mathcal{Q}$-value for every state-action pair. Furthermore, as DPGD only works in continuous action space, we embed a k-nearest neighbor algorithm into DQL to quickly find a valid action in the discrete space nearest to the DPGD output. Finally, we verify the faster convergence of the DNAF-based DQL through extensive simulations.

cs.LG↗

arXiv2019

Multi-Tenant Cross-Slice Resource Orchestration: A Deep Reinforcement Learning Approach

With the cellular networks becoming increasingly agile, a major challenge lies in how to support diverse services for mobile users (MUs) over a common physical network infrastructure. Network slicing is a promising solution to tailor the network to match such service requests. This paper considers a system with radio access network (RAN)-only slicing, where the physical infrastructure is split into slices providing computation and communication functionalities. A limited number of channels are auctioned across scheduling slots to MUs of multiple service providers (SPs) (i.e., the tenants). Each SP behaves selfishly to maximize the expected long-term payoff from the competition with other SPs for the orchestration of channels, which provides its MUs with the opportunities to access the computation and communication slices. This problem is modelled as a stochastic game, in which the decision makings of a SP depend on the global network dynamics as well as the joint control policy of all SPs. To approximate the Nash equilibrium solutions, we first construct an abstract stochastic game with the local conjectures of channel auction among the SPs. We then linearly decompose the per-SP Markov decision process to simplify the decision makings at a SP and derive an online scheme based on deep reinforcement learning to approach the optimal abstract control policies. Numerical experiments show significant performance gains from our scheme.

cs.NI↗

arXiv2019

Evaluation Mechanism of Collective Intelligence for Heterogeneous Agents Group

Collective intelligence is manifested when multiple agents coherently work in observation, interaction, decision-making and action. In this paper, we define and quantify the intelligence level of heterogeneous agents group with the improved Anytime Universal Intelligence Test(AUIT), based on an extension of the existing evaluation of homogeneous agents group. The relationship of intelligence level with agents composition, group size, spatial complexity and testing time is analyzed. The intelligence level of heterogeneous agents groups is compared with the homogeneous ones to analyze the effects of heterogeneity on collective intelligence. Our work will help to understand the essence of collective intelligence more deeply and reveal the effect of various key factors on group intelligence level.

cs.AI↗