Source-linked AI summary
Game-theoretic Resource Allocation Methods for Device-to-Device (D2D) Communication
Lingyang Song, Dusit Niyato, Zhu Han, Ekram Hossain
TL;DR
D2D networks must allocate shared radio resources among cellular and D2D users with competing or cooperative interests. This article reviews game-theoretic models for distributed D2D resource allocation, including direct communication and D2D LANs, and outlines open research directions. It concludes that game models support stable and efficient allocation, with cooperative models applicable when D2D LAN collaboration is required.
Problem
Cellular and D2D users must solve distributed resource-allocation decisions while sharing spectrum and managing interference.
Method
The article reviews game-theoretic models for D2D resource allocation, including auction, coalitional, and cooperative game approaches.
Results
Game models are useful for designing radio-resource allocation algorithms that achieve stable and efficient outcomes for D2D communication.
Takeaways & Limitations
Cooperative game models such as coalition formation games can support D2D LANs when collaboration among mobiles is required.
Abstract
from arXiv · showhide
Device-to-device (D2D) communication underlaying cellular networks allows mobile devices such as smartphones and tablets to use the licensed spectrum allocated to cellular services for direct peer-to-peer transmission. D2D communication can use either one-hop transmission (i.e., in D2D direct communication) or multi-hop cluster-based transmission (i.e., in D2D local area networks). The D2D devices can compete or cooperate with each other to reuse the radio resources in D2D networks. Therefore, resource allocation and access for D2D communication can be treated as games. The theories behind these games provide a variety of mathematical tools to effectively model and analyze the individual or group behaviors of D2D users. In addition, game models can provide distributed solutions to the resource allocation problems for D2D communication. The aim of this article is to demonstrate the applications of game-theoretic models to study the radio resource allocation issues in D2D communication. The article also outlines several key open research directions.
I. INTRODUCTION
D2D communication lets nearby devices reuse cellular spectrum through direct or multi-hop links, but this creates interference and distributed resource-allocation challenges. The article presents game theory as a framework for modeling these interactions and developing distributed allocation solutions.
- D2D Communication Models: D2D communication supports one-hop direct links and multi-hop, cluster-based D2D LAN communication over licensed spectrum.
- Benefits and Challenges: Underlay spectrum reuse can improve coverage, offloading, energy efficiency, throughput, and spectrum efficiency, while also causing interference to cellular users.
- Benefits and Challenges: Interference management and coordination must protect cellular users while meeting D2D users’ QoS requirements and maximizing spectrum utilization.
- Need for Distributed Solutions: Existing optimization-based approaches require users to share a single objective and be fully controlled through extensive signaling.
- Need for Distributed Solutions: Cellular and D2D users may pursue self-interested benefits through cooperation or competition, creating distributed decision problems involving users and the base station.
- Game-Theoretic Approach: Game theory provides tools for analyzing interdependent rational players and designing channel-assignment, power-control, and cooperation-enforcement solutions.
- Article Scope: The article reviews game-theoretic resource-allocation approaches for D2D communication and identifies open research directions for other D2D problems.
II. BASICS OF D2D COMMUNICATION
D2D communication supports direct and local-area-network scenarios, with overlay or underlay spectrum access and centralized or distributed resource allocation. Underlay reuse improves spectrum efficiency but requires careful interference management and coordination among cellular and D2D users.
- D2D communication scenarios: D2D communication includes direct one-hop links and local-area-network scenarios supporting group communication, relaying, and collaborative sensing.D2D LANs can distribute popular content through selected seed UEs, relay files beyond eNB coverage, or aggregate sensed data at sink UEs.
- Spectrum access: Overlay sharing separates cellular and D2D spectrum, whereas underlay sharing allows simultaneous reuse of the same spectrum.Overlay reduces intra-cell interference but limits flexible spectrum access; underlay improves utilization and efficiency while requiring interference control.
- Spectrum access: Underlay operation can improve system performance through frequency reuse but requires greater signaling overhead and control.The article focuses on a two-tier network containing cellular and D2D users, with interference both across and within tiers.
- Resource allocation: Interference avoidance, randomization, or cancellation must reduce dead zones while balancing the interests of operators, cellular users, and D2D users.Game theory and auction theory provide distributed or mathematical tools for channel assignment, power control, cooperation, and direct-communication resource allocation.
- Resource allocation: D2D resource allocation assigns D2D users to resource blocks while accounting for co-channel interference from cellular users.Local allocation adjusts D2D resources around fixed cellular allocations, while global allocation controls resources for both user types.
- D2D LAN scenarios: D2D LAN cluster formation can be formulated as a cooperative game with a distributed coalition-formation algorithm converging to a stable solution.This clustering supports network-controlled group communication and other multi-user LAN functions.
III. GAME THEORY-BASED RESOURCE ALLOCATION FOR D2D DIRECT COMMUNICATION
The article applies game-theoretic models to D2D direct-communication resource allocation and illustrates an auction-based approach. These models address radio-resource allocation problems through game-based analysis.
- Game-theoretic allocation: Game-theoretic models are used to solve resource-allocation problems for D2D direct communication.The section specifically discusses different game models and an auction-based resource-allocation model.
A. Different Game Models for D2D Direct Communication •
The article surveys noncooperative, Stackelberg, and auction game models for D2D direct communication. These approaches trade off distributed simplicity, signaling overhead, and allocation performance.
- Noncooperative power control: A noncooperative power-control game models D2D transmitters choosing channel transmit powers to minimize their total transmit power.Transmit powers are updated iteratively in a self-organizing manner while multiple D2D users reuse cellular spectrum.
- Stackelberg game: A Stackelberg game treats cellular users as leaders and D2D users as followers in channel sharing.Leaders set access fees, while followers choose transmit power and channels to maximize throughput minus access cost; an equilibrium can be reached.
- Auction models: Combinatorial auctions let bidders bid on resource packages rather than individual resource blocks, encouraging fuller preference expression.Iterative auctions update prices according to demand and supply, using pricing to achieve efficient allocation rather than as the primary objective.
- Auction models: Auction games can allocate resources among D2D links reusing cellular channels through iterative bids and provisional allocations.The article presents a combinatorial-auction-based approach for D2D direct communication.
- Approach trade-offs: Noncooperative games typically provide suboptimal solutions with smaller signaling overhead, whereas auctions can achieve near-optimal performance with signaling overhead.The comparison summarizes the principal trade-off among the reviewed game-theoretic approaches.
B. Combinatorial Auction-Based Resource Allocation
The proposed reverse iterative combinatorial auction allocates cellular resource blocks to D2D users through package bidding and iterative price updates. Its performance improves with more D2D pairs before saturation, while random allocation performs worse than cellular-only operation.
- Auction design: The auction treats cellular and D2D users as auction participants and defines utility as allocated-channel sum-rate minus signaling-overhead cost.Cellular networks act as bidders for packages of resource blocks, and the auction outcome assigns a corresponding package to each bidder.
- Auction design: D2D users bid on combinations of resource blocks, while auctioneer prices are repeatedly raised or reduced according to demand and supply.The process continues through multiple iterations until links are auctioned or channels receive packages and an equilibrium solution is reached.
- Auction properties: The combinatorial auction has finite convergence, price monotonicity associated with efficiency improvement, and complexity O(n (2m −1) + t).Here, m is the number of items, n the number of bidders, and t the total number of iterations.
- Simulation results: R-I-CA system sum-rate first increases and then saturates as the number of D2D pairs increases.The passage attributes saturation to increased co-tier interference, which makes further resource purchases unaffordable for bidders.
- Simulation results: Random allocation performs worse than assigning all mobiles to traditional cellular mode in the reported comparison.The cellular-only case connects all mobiles to the eNB and may experience both co-tier and cross-tier interference.
IV. GAME THEORY MODELS FOR D2D LOCAL AREA NETWORKS
Game-theoretic models frame D2D LAN resource allocation as cooperative or noncooperative interaction among users. Cooperative models coordinate users toward shared utility and can improve network performance, while applications include group and multi-hop communication.
- D2D LAN resource allocation for group and multi-hop relay communication can be modeled using game-theoretic approaches.
- Noncooperative users make individual decisions that may cause severe interference, whereas cooperation coordinates behavior toward a common utility objective.
- Cooperative approaches can achieve better network performance than noncooperative approaches by having mobiles cooperate to maximize utility.
- Cooperative game models include coalitional and Nash bargaining games for coordinating strategies and dividing coalition payoffs.
- Coalition-based approaches are applied to group communication and content distribution in D2D networks.
A. Coalition Formation Game Models for D2D Communication
The paper surveys coalition formation models for D2D communication, covering coalition structures, payoff types, convergence mechanisms, relaying, sensing, and socially informed resource allocation.
- Coalitions represent agreements among D2D players to act as a single entity, with values depending on coalition members alone or on the surrounding partition.
- Coalitional games may use transferable payoffs, where utility is distributable, or nontransferable payoffs, where utilities cannot be allocated arbitrarily.
- The merge-and-split algorithm merges mutually beneficial coalitions and splits coalitions when splitting improves payoffs.
- The merge-and-split algorithm converges to stable coalitions in which no player prefers a merge or split yielding higher utility.
- Coalition graph games model multi-hop D2D relaying through graph-dependent utilities, directed trees, route selection, and power-level selection.
- Overlapping coalition formation supports collaborative smartphone sensing by allowing users to contribute resources to multiple sensing-task coalitions.
- Matching and contract theory can jointly optimize social-network and D2D-LAN layers for resource allocation in mobile social networks.
B. Coalitional Game-Based Resource Allocation for Efficient Content Distribution
For D2D content distribution, the paper uses coalition formation to organize cellular and D2D users sharing radio resources. Switch-based coalition adjustments converge to stable groupings, and the proposed approach outperforms a noncooperative one in service rate.
- A content-distribution scenario serves N users requesting the same file, with K seeds distributing it to N − K normal UEs through D2D links.
- The resource-allocation problem depends on selecting resource blocks and determining which D2D links share them.
- Each coalition contains one cellular UE and D2D UEs using the same resource block, with coalition value defined as the sum rate of cellular and D2D links.
- Because different coalitions use different resource blocks, they do not interfere; a switch moves a player when the two coalitions’ total value strictly increases.
- The switch-based process can reach stable coalitions in which no switch operation is preferred.
- The proposed coalition game-theoretic approach performs much better than the noncooperative approach in cumulative service rate.
V. POTENTIAL RESEARCH DIRECTIONS IN DESIGNING GAME THEORETIC METHODS FOR D2D COMMUNICATION
The paper identifies game-theoretic tools as applicable to many additional D2D resource-allocation problems and outlines potential research directions.
- Game-theoretic tools can address many other resource-allocation problems for D2D communication.
- The paper outlines several potential research directions for designing game-theoretic methods for D2D communication.
A. Device-to-Device Direct Communication •
D2D direct communication requires distributed mode selection and coordination across users, cells, and resource dimensions. The article maps these challenges to evolutionary, cooperative, multi-level, and differential game models.
- Multi-cell D2D operation requires coordination among base stations because D2D can reduce transmission time slots while creating intercell interference concerns.
- Cooperative game models can support multi-cell optimization and intercell interference coordination, while signalling exchange introduces additional time-delay cost.
- Users choose between cellular and direct D2D modes based on distance, eNB transmission quality, interference, and performance requirements.
- Distributed transmission-mode selection is required for scalable D2D-network implementation, with evolutionary games proposed to balance cellular and D2D traffic.
- Joint allocation of space, time, frequency, power, and devices motivates multi-level games, combining noncooperative power allocation with auction-based resource assignment.
- Differential games can analyze and design hybrid networks by studying the tradeoff between total transmit power and system throughput.
B. Device-to-Device Local Area Networks •
D2D local area networks depend on cooperation, incentives, and matching among mobiles. The article presents cooperative and coalitional models for resource allocation, alongside incentive and crowdsourcing applications.
- Helper relays can extend coverage or improve spatial diversity, so rewarding systems and contract games can motivate intermediate-node participation.
- Matching theory can assign mobiles to interested targets and coordinate resource allocation for multi-file sharing and collaborative indoor localization or navigation.
- These matching-based mechanisms can generate cost-effective alternatives to expensive centralized solutions.
- The article categorizes game models by scenario: noncooperative and auction games suit D2D direct communication, whereas LANs use cooperative games.
- Cooperative D2D LANs require mobiles to collaborate, motivating cooperative models such as coalition formation games.
- The article presents auction and coalitional models and outlines game-theoretic research directions for important D2D radio resource-management problems.