2021 Blockchain incentivized data forwarding in MANETs - Strategies and challenges

Last updated Dec 11, 2022

• Source zotero journal article
• Type subject
• Status needs revision
• On networks manet
• On blockchain

Metadata

• CiteKey:: machadoBlockchainIncentivizedData2021
• Type:: journalArticle
• Author:: Caciano Machado Carlos Becker Westphall Carla Merkle Westphall
• Editor::
• Publisher::
• Series::
• Series Number::
• Journal::
• Volume:: 110
• Issue::
• Pages::
• Year:: 2021
• DOI:: 10.1016/j.adhoc.2020.102321
• ISSN::
• ISBN::
• Format:: PDF

Abstract

Abstract Recently, blockchain trustless properties started to be investigated to design cooperation enforcement mechanisms in many systems. This paper presents a comprehensive and detailed review of works on blockchain-enabled data forwarding incentives for multi-hop MANETs. We contextualize the problem of selfish misbehavior in networks composed of routers that are property of different participants: community, D2D, and vehicular networks, including DTN alternatives. We discuss how uncooperative behavior from multiple device owners leads to unreliable communication affecting trust in MANETs. We summarize pre-blockchain incentive mechanisms for data forwarding, classified as credit-based and reputation-based, and outline game-theoretic approaches. We discuss blockchain features useful for data forwarding incentives in multi-hop MANETs, detailing off-chain mechanisms that have been applied in the state-of-the-art. We describe the critical points in the state-of-the-art based on research papers, patents, and products. Finally, we discuss and summarize existing strategies and challenges for further research.

Files and Links

• Url::
• Uri:: http://zotero.org/users/5055703/items/9E5EDXJS
• File:: machado_et_al_2021_blockchain_incentivized_data_forwarding_in_manets.pdf
• Local Library:: machado_et_al_2021_blockchain_incentivized_data_forwarding_in_manets.pdf

Tags and Collections

• Keywords::
• Collections:: PoN

• ["] Specific mobile ad hoc networks (MANET) have been proposed to expand network coverage to regions that conventional networks cannot reach. Page 1
• ["] MANETs require cooperative sharing of resources to enable reliable data forwarding. However, misbehaved nodes could undermine network reliability by acting selfishly, taking advantage of cooperation from other devices, and avoiding making their resources available. This behavior is also known as free-riding. Page 1
• ["] Reputationbased mechanisms are prone to second-order free-riding [8], which consists of devices that avoid contributing to the reputation mechanism. Page 1
• ["] Recently, blockchains started to be adopted to provide financial compensation for collaborative participants in MANETs. Blockchains have trustless properties Page 1
• ["] Misbehavior in MANETs can be classified according to the intention of the participants. Unintentional misbehavior appears independently of a participant’s will, such as node mobility and transmission errors. Intentional misbehavior can be subdivided into malicious and selfish. Malicious misbehavior consists of attacks such as vandalism, denial of service, and exploration of protocol vulnerabilities. Selfish misbehavior consists in refusal to cooperate in the network operation providing computational, network and energetic resources because this implies an opportunity cost. Page 2
• ["] The most direct effects of selfish misbehavior in the reliability of packet-switched MANETs are packet loss and delivery delays. These effects are due to selfish routers that drop or delay packets, and avoid to cooperate, overloading other cooperative routers. Page 2
• ["] Disputes for channel allocation in unlicensed radio frequencies is an example of selfish misbehavior in the physical layer Page 2
• ["] Multi-hop networks composed of routers that are property of different owners are prone to the free-riding problem. In this context, a free-rider is a router that consumes more than contributes to the network, i.e., produces messages that are rightly forwarded by other routers, but do not relay messages from other devices reciprocally. Page 3
• ["] In a general definition derived from social sciences, trust is the degree of subjective belief about a particular entity’s behavior. In this context, the reputation of an entity is established from its previously performed actions. Page 3
• ["] trust is the belief that an entity is capable of performing reliably, dependably, and securely in a particular case. Page 3
• ["] Trust management systems intend to improve network reliability and usually are based on monitoring, directly and indirectly, nodes’ behavior. The rationale is to trust on the most reliable nodes, i.e., nodes with a higher probability of forwarding packets or contributing to the routing protocol. Page 3
• ["] selfish users tend to give preference to their own traffic in detriment of others’ traffic Page 3
• ["] Selfish users avoid sharing capital expenditure (CAPEX: routers, antennas, cabling, and licensing) and operational expenditure (OPEX: backhaul contracts, electric energy, working hours, and maintenance) to forward traffic that is not useful to them. Page 3
• ["] In contrast, selfish misbehavior does not affect networks composed of routers that are property of institutions such as universities, companies, and government offices. In such situations, network devices are configured by the same authority and do not present the free-riding behavior in data forwarding and routing protocols. Page 4
• ["] Community networks [1–3] are a type of wireless mesh network [25] that aims to provide last-mile infrastructure for Internet access in underserved or underdeveloped areas….Community networks are composed of shared low-cost off-the-shelf routers that form a cooperative multi-hop network that interconnects user devices to an ISP. These networks are more static than typical MANETs and can also adopt wired communication between hops. Compared to other MANETs, trust in community networks is easier to achieve due to more fixed and stable nodes. Page 4
• ["] In contrast to community networks, device-to-device (D2D) networks [26] (or even client wireless mesh networks) [27] are inherently more dynamic and ephemeral than community networks. Routers are personal mobile devices such as notebooks, smartphones, tablets, and other devices with WiFi, Bluetooth, NFC, or other short-range wireless technology. Nodes can join and leave the network using and contributing to the operation of the network while active….It is harder to establish trust between nodes because the neighborhood is constantly changing, and reputation information becomes obsolete quickly. Page 4
• ["] A vehicular ad hoc network (VANET) [5] is a MANET type that vehicles work as routers that relay data. Usually, they serve for vehicle coordination, traffic information, and road services (emergency, gas stations, restaurants), but they can also provide last-mile Internet access as in other wireless mesh networks….VANETs also present similar challenges for managing trust as in D2D networks Page 4
• ["] Delay-tolerant networks (DTN) [6] are also known as disruptiontolerant networks or opportunistic networks. The objective of DTNs is to enable communication in MANETs that are constantly partitioned. This type of network follows the store-carry-forward paradigm…Instead of immediately forwarding received data, routers carry data until they find an opportunity to forward them to another router toward the destination. This process increases latency and jitter significantly but enables communication for classes of applications that are not sensitive to such delays. Page 4

Page 5

• ["] An incentive mechanism can be defined as a system rule whose goal is to induce participants to act in a specific way. Collaboration could be achieved with rewards to stimulate cooperation or punishments to discourage misbehavior. Page 5
• ["] Incentive mechanisms assume that participants act rationally, from an economic perspective. Page 5
• ["] incentive mechanisms do not necessarily provide strong authentication of entities. Instead, they contribute to identifying the trustworthiness of peers and enforce cooperation using mutual incentives Page 5
• ["] Credit-based mechanisms, illustrated in Fig. 8, model the data-forwarding task as a service that can be valuated and charged. These models incorporate a form of virtual currency to regulate the dealings between the various nodes for data forwarding in multi-hop networks. Virtual currency is used by source and destination nodes to pay forwarder nodes. Also, forwarder nodes are incentivized to relay messages to earn credit because they need credit when they assume the role of source or destination of messages to pay other forwarder nodes. Page 5
• ["] Two approaches have been widely adopted for secure credit-based mechanisms: tamper-resistant hardware that secures credit accounting with dedicated hardware modules attached to the network interfaces; virtual banks that depend on a trusted third-party service responsible for centralized accounting. Page 5
• ["] Furthermore, credit-based mechanisms present reciprocity limitation, i.e., a participant credit is bound to its contribution in data forwarding. Consequently, if there is no traffic demand to be forwarded by other participants, a high demanding node cannot acquire enough credits to pay other nodes to forward its data. Some credit-based mechanisms solve this issue by introducing an external currency Page 5

Page 6

• ["] Bogliolo et al. [10] were the first to suggest using blockchains for credit-based incentives in MANETs in order to eliminate the need for trusted third-parties or tamper-resistant hardware. Page 6
• ["] Reputation-based mechanisms [20–23] evaluate the reputation of nodes to forward packets through the most reliable nodes. The reputation of a node increases when it carries out rightly the task of forwarding data sent by its neighbors. Page 6
• ["] ICARUS [48] is an example of a hybrid incentive mechanism that combines reputation and credit techniques. Page 6
• ["] A well-known challenge of reputation mechanisms is the secondorder free-riding problem [8], i.e., participants who do not spend resources detecting and punishing free-riders. Page 6
• ["] Game theory [54] models situations in which multiple participants select strategies that have mutual consequences. Page 6
• ["] If players select strategies such that no one can unilaterally change its strategy to gain more payoff, we say that the game reaches a Nash equilibrium. Page 6
• ["] Algorithmic game theory design [55] is a subarea of game theory that deals with the design of games. It studies optimization problems where the underlying data is a priori unknown to the algorithm designer and must be, implicitly or explicitly, extracted from selfish participants, e.g., via a bid. The high-level goal is to design a protocol, i.e., an incentive or cooperation enforcement mechanism, that interacts with participants so that even selfish non-cooperative behavior yields a desirable outcome. Page 6
• ["] The book Game Theory in Wireless and Communication Networks [56] presents a comprehensive compilation of game-theoretic works for multi-hop MANETs Page 6
• ["] Besides the potential of deploying sustainable networks in areas not covered by conventional services, incentivized MANETs could also minimize nonessential and undesired traffic. Page 7
• ["] Blockchains are distributed databases organized as sequential chains of blocks that store transactions Page 7
• ["] Each transaction in Bitcoin, say Alice paying Bob 10 BTC, has one or more transaction outputs (TXO), which serve as sums of spendable BTC. These unspent sums are called unspent transaction outputs (UTXO). They remain UTXOs until the owner (Bob, for example) redeems them to pay someone else. After that, they are referred to as spent TXOs. Page 7
• ["] In a UTXO based blockchain, there are no accounts or wallets at the protocol layer. Instead, coins are stored as a list of unspent transaction outputs or UTXOs. Page 7
• ["] Rather than following in Bitcoin’s principles, smart contractbased blockchains have chosen to employ an account strategy. Instead of having each coin uniquely referenced, coins are represented as a balance within an account. Accounts can either be controlled by a private key or a smart contract. Page 7
• ["] Consensus algorithms are the core of blockchains and serve to establish agreement of the content and ordering of transactions among nodes Page 7
• ["] The trustless property comes from the possibility of any node validating the integrity of blocks independently, eliminating the need for prior trust establishment with third-parties. Page 7
• ["] Blockchain can be classified according to the participation of nodes. Zheng et al. [71] taxonomy characterizes blockchains as public when they are permissionless, i.e., they allow any node in the world to participate in the consensus protocol. In opposition, consortium or private blockchains are permissioned, i.e., they require authentication of nodes. Page 7
• ["] Blockchains systems present scalability issues that limit their practical use for many applications: low throughput, that can be expressed in transactions per second (tps) capacity; high financial costs, that involves the amount of cryptocurrency spent in transaction fees; and high storage costs, that relates to the size of the blockchain file. Transaction confirmation time is another issue related to the frequency that new blocks are created. Page 7
• ["] On-chain approaches try to change aspects from consensus algorithms to achieve slightly better scalability Page 7
• ["] Off-chain mechanisms are decoupled from the main chain consensus protocol and could achieve high scalability improvements. Page 7

Page 8

• ["] Blockchain channels have been proposed to enable the secure exchange of transactions among parties outside of the blockchain (named off-chain). These transactions act as an escrow or promissory notes and are settled later on the blockchain. A channel represents the relationship among parties, outside of the blockchain, and can be classified as micropayment channels and state channels. The former represents mechanisms that serve only for cryptocurrency transfer transactions [84,85], and the latter is a generalization that intends to support state transitions for smart contracts Page 8
• ["] The creation of a micropayment channel requires a blockchain transaction, called funding transaction, and involves regular transaction times and cryptocurrency fees costs Page 8
• ["] Parties exchange signed transactions, called commitment transactions, that alter the initial balance value Page 8
• ["] The settlement transaction represents the final state of the channel and is settled on the blockchain by any of the parties Page 8
• ["] In the end, only two transactions are recorded on the blockchain: the funding transaction that established the channel and a settlement transaction that allocated the final balance correctly between the participants. The settlement transaction must be signed among parties Page 8
• ["] There are many problems to maintain micropayment channels correct operation. If one party disappears before establishing at least one commitment transaction, the other party will lose the funds, since there is no commitment performed. Besides, if one party broadcasts a commitment transaction in his/her favor, the other party could be wrongly paid for a service provided. Page 8
• ["] A sidechain is a blockchain with its own independently secured consensus algorithm and is pegged to another blockchain. Value can be transferred from one blockchain to another by relaying simple payment verification (SPV) [67] proofs. Page 8
• ["] Disputes that happen due to fraud in the leaves are escalated to parentchains towards the mainchain. Thus, the mainchain acts as the final arbitrator in the case of unresolved disputes. Page 8
• ["] MANETs are also starting to adopt cryptocurrencies to improve network connectivity trust without the need for trust in third-parties (trustless). Page 9
• ["] Typically, credit-based incentives are limited by reciprocity, i.e., a participant cannot use network services beyond its contribution. This limitation does not exist with financially incentivized MANETs with cryptocurrencies. Any participant can use network services, given that they have enough cryptocurrency to pay for it. Page 10
• ["] ERC-20 [127] is a protocol standard that defines smart contract rules for issuing tokens on Ethereum. Page 10
• ["] Permissioned blockchains, adopted by some systems [124,128], allow better performance, though they are not trustless. Page 10
• ["] many MANET systems adopted off-chain mechanisms, such as channels and childchains, to enable faster and cheaper transactions. Page 10
• ["] Micropayments channels (Section 4.1) have been widely applied in incentivized MANETs for implementing ideas similar to conventional escrows and checks. For instance, establishing a channel needs an escrow that the parties should deposit as collateral for transactions. Likewise, micropayments can be compared as checks, because the transaction information is kept off-chain by the parties until settlement in the main chain. Page 10
• ["] Childchains have a few proposals for payments in incentivized MANETS but no public implementation yet. Page 10 ^87949e
• ["] Both channels and childchains are deployed over smart contracts Page 10
• ["] MeshDapp [124] deploys smart contracts to estimate demand and supply of network forwarding services, and define prices based on estimations. Page 10
• ["] Time-locked puzzles are mechanisms that hide information in the blockchain for a specific time or until certain conditions are satisfied. Bitcoin’s time-locked puzzle implementation allows us to retain a reward until one of three conditions is met: until a specific time passes; until a node solves the puzzle; or until the solution is revealed. Page 10
• ["] The double incentive forwarding makes the forwarders lose their reward unless they forward the data intact to the next-hop as soon as possible and creates an incentive for assisting other forwarders. Page 10
• ["] The all or nothing strategy pays the reward to all nodes only after the data is delivered to the destination. Instead of distributing secrets in advance, the final receiver acknowledges the data delivery to the sender, and the sender then unlocks the puzzles for all the forwarders. Page 10

Page 11

• ["] The contract forwarding works without prior establishment of the forwarding path as the two previous strategies. The sender negotiates a forwarding contract with another node to bring the data to the recipient. It is then up to the node that accepted the forwarding contract to deliver the data as fast as possible. Page 11
• ["] RouteBazaar [111] uses blockchains to build trust between Internet autonomous systems (AS)….Agreements are registered in the blockchain and identified by an anonymous tag created by parties. The forwarding proof for a specific pathlet is also written in the blockchain….Clients register payment proof directly in the blockchain too. Page 11
• ["] RouteBazaar has no implementations, and the need to write in the blockchain so often would cause considerable overhead. Page 11
• ["] Althea aims to incentivize communities to deploy last-mile connectivity to the ISP. It uses the Babel protocol [130] to determine the routes of infrastructured mesh networks. The routing protocol also incorporates price metrics that consider how much each router owner wants to receive as payment for data forwarded and mechanisms to verify announced metrics. Thus, routes are determined according to traditional cost metrics and the proposed price metrics. The weight of price metrics is adjustable so that users can define their link preference between price and quality. Page 12
• ["] The forwarding incentive scheme relies on payment for forwarded data using micropayment channels. The current version of Althea uses the Ethereum blockchain with a low-overhead micropayment channel mechanism called Guac [131]. Each node that wants to have data forwarded establishes Guac micropayment channels and VPN tunnels with its neighbors for payment. VPN tunnels serve as an accounting mechanism to control data delivery with neighbors. Moreover, nodes pay neighbors only after their data is forwarded, and forwarders can block or shape the traffic of bad payers. VPN accounting also serves as a reputation mechanism to avoid nodes that provide low-quality service. Additionally, Althea nodes also create VPN tunnels with exit nodes Page 12
• ["] Rightmesh [118] proposes to incentivize D2D DTNs with Ethereum micropayments. It has an Android API to build applications using a proprietary protocol stack that operates over Bluetooth and WiFi technology. Page 12
• ["] LOT49 [117] proposes D2D networks incentivized with Bitcoin payments using the Lightning Protocol for channel micropayments. They also propose a new scheme for aggregated signatures [87,133] in micropayment channels to minimize the incentive protocol overhead and increase the bandwidth available for data delivery. Page 12
• ["] Like Althea, AMMBR [128] aims to disseminate the Internet with blockchain incentivized community networks in the last mile to the ISP. AMMBR supports the mesh routing protocol BATMAN-Adv [135] and proposes the development of a new one based on BMX7 [136]. AMMBR launched its cryptocurrency (AMR) and designed its router hardware. The proposed router is modular and extensible, supporting modules for blockchain mining, multiple radio technologies, and IoT-related features. Page 13
• ["] Trautmann and Burnell’s patent [113] describes a system that introduces a Proof-of-Routing scheme that can securely implement a blockchain network and provide useful consensus. Their blockchainbased router idea includes different nodes that process data packets between endpoints. Nodes can include router nodes, which analyze and route data packets, and block nodes that manage collections of specially labeled packets and generate new blocks in the blockchain….This mechanism incentivizes data packets to be signed and forwarded to their respective destinations and stimulates router nodes to not adhere to free-riding behavior. Page 13
• ["] MeshDapp [124,138] focuses on balancing mesh network service costs (CAPEX and OPEX) and respective payments to enable sustainable network infrastructure. Their approach uses Ethereum smart contracts to automate fair accounting and money transfers for the network service provided. Page 13

Page 14

Page 14

• ["] Intermittent and low-bandwidth connectivity can affect MANETs, mainly DTNs. Consequently, devices could stay out of sync with the blockchain for long periods and unable to perform on-chain transactions. Off-chain mechanisms enable transactions to be performed securely between nodes in a MANET, even when the network is partitioned and without Internet connectivity. Page 14
• ["] Both proxy-based and key issuer approaches characterize dependence on trusted third-parties that eliminate the trustless property. Page 14
• ["] Besides payment proofs, blockchain incentivized MANETs need forwarding proofs that confirm that nodes contributed to data forwarding to enable fair payments and prevent undue billing. Page 14
• ["] We divide forwarding proofs from the state-of-the-art according to two criteria: mechanisms (Table 4) and trust (Table 5). Page 14
• ["] Mechanisms define how forwarding proofs are implemented. On the one hand, monitoring mechanisms implement traffic metering in the MANET with proxies and tunnels. On the other hand, receipt mechanisms consist of packet delivery acknowledgments with piggybacked receipts. Page 14

Page 15

• ["] The ownership of receipts is enough for forwarding proofs because the destination already acknowledged the packet’s delivery. Page 15 Page 15
• ["] In the trust criteria, we divide forwarding proofs in trusted third-parties and trustless approaches. Trusted third-parties approaches assume trust in specific elements of the network architecture to ensure forwarding proofs, despite trustless payment mechanisms. Page 15
• ["] We believe that efficient solutions for trustless forwarding proofs should rely on algorithmic game theory design to model incentives for distributed accounting. The idea presented in the patent from Trautmann and Burnell [113] seems to follow this approach, though it needs further investigation. Page 15 ^6c1a5c
• ["] Most systems from the state-of-the-art implement forwarding incentives on top of existing routing paths. Additionally, some systems are routing protocol agnostic Page 15
• ["] Besides forwarding incentives, price-aware routing protocols have been explored in some works [111,112]. For example, Althea incorporates price costs in link metrics in addition to typical costs such as link speed and quality of service, as illustrated in Fig. 13. As a result, paths are determined on a market basis where the price of the links is taken into account. Page 15

Page 15

• ["] Some works suggest using network service provided as a basis for proofs in blockchain consensus to produce (mine) new cryptocurrency [113,116,120,128]. For example, a router that proves that it contributed to traffic forwarding or the convergence of a routing protocol could receive cryptocurrency as a Proof-of-Networking (PoN) [145] reward likewise Bitcoin’s Proof-of-Work. The idea that is closer to PoN has been proposed by Trautmann and Burnell [113] in their Proof-of-Routing scheme. Page 15 ^ab8b1f
• ["] One problem that is barely discussed in the state-of-the-art is how to deal with different network quality of service (QoS) requirements. Real-time audio and video communication, online services, and delay-tolerant applications have distinct network requirements in terms of bandwidth, latency, and jitter. Systems should deploy some sort of network resource reservation and queueing policies to accomplish strict network requirements. At least, the system should allow clients to detect whether services are being provided as advertised or not. These features would require more complex forwarding proof mechanisms. Page 15 ^3fb577
• ["] Incentive mechanisms in MANETs could leak information about localization and trajectory of nodes and users when they create public on-chain transactions for payments. This security issue could also inhibit users from using MANET incentivized services. Page 15 ^29ba0b

Page 16