Search

CN-122027666-A - Internet of vehicles road condition information consensus system and route planning method based on consensus system

CN122027666ACN 122027666 ACN122027666 ACN 122027666ACN-122027666-A

Abstract

The invention relates to a road condition information consensus system of the Internet of vehicles and a route planning method based on the consensus system, wherein the system comprises a dynamic consensus group screening module for selecting a coordinator from edge nodes and generating a consensus group through dynamic election, a first-stage competition block out module for competing out blocks and reporting road condition information rights according to effective capacity and credit, a second-stage competition block out module for competing out blocks and road condition reporting rights again after excluding first-stage winners, a double-stage differential rewarding module for giving different credit rewards to winners of the first-stage competition block out from the second-stage competition block out, and a credit management and self-healing module for managing credit values of edge nodes of the whole network and realizing fault recovery based on established double-layer blacklist-self-healing logic.

Inventors

  • WANG SHAOQIANG
  • ZHENG JINTAO
  • KANG LEI
  • Qian Jiapeng
  • Min Haojia

Assignees

  • 长春大学

Dates

Publication Date
20260512
Application Date
20260410

Claims (10)

  1. 1. The system is characterized by comprising a dynamic consensus group screening module, a first-stage competition out block module, a second-stage competition out block module, a double-stage differential rewarding module and a reputation management and self-healing module, wherein the dynamic consensus group screening module is used for selecting a coordinator of a consensus process from edge nodes, the coordinator is a winner or an initial node of the last round, and generates a consensus group through dynamic election, during the dynamic election, the edge nodes with the top 50% of comprehensive grading are firstly used as candidate pools, and then logarithmic constraint is carried out according to the total number N of the edge nodes of the network to determine the number K of the consensus group members; The first stage competition block-out module is used for competing out blocks and reporting road condition information rights according to the effective capacity and the credit; The second stage competition block out module is used for carrying out block out and road condition reporting right competition again after the winner of the first stage is eliminated; The two-stage differentiated rewarding module is used for giving different credit rewards to winners of the first-stage competition block-out module and the second-stage competition block-out module, giving fixed basic rewards to the winners of the first stage and giving capability matching dynamic rewards to the winners of the second stage; The reputation management and self-healing module is used for managing reputation values of the edge nodes of the whole network and realizing fault recovery based on an established double-layer blacklist mechanism, global big privilege and self-healing logic.
  2. 2. The internet of vehicles traffic information consensus system according to claim 1, wherein the process of dynamically electing to generate the consensus group by the dynamic consensus group screening module is: Step A1. Start and mortgage: the coordinator broadcasts a message to open a mortgage window when each round starts, and the edge nodes participating in consensus of the whole network calculate self comprehensive scores, submit the comprehensive scores of the capacity and the credit to the coordinator and simultaneously submit a mortgage space; the expression of the comprehensive score is ComScore =c×r; Wherein ComScore is the composite score, C is the physical capacity, R is the reputation value; Step B2, top50% sequencing and screening: the coordinator collects the submitted information of all qualified edge nodes, sorts the submitted information from high to low according to scores, selects the edge nodes with the top 50% of the ranks as candidate pools, carries out logarithmic constraint according to the number N of the total edge nodes of the network, and determines the number of members of the consensus group And then selecting the first K members from the candidate pool to form a consensus group, and broadcasting the consensus group in the whole network.
  3. 3. The internet of vehicles traffic information consensus system according to claim 1 wherein the first stage contending block module contends for a block by: a2, calculating effective capacity based on the physical capacity C, wherein the effective capacity E and the physical capacity C are in a nonlinear power function relation; step B2, performing capacity map traversal and deadline calculation: The edge node generates a challenge value based on the last block hash, performs traversal search in a locally stored capacity graph, searches for an optimal random number matched with the challenge value, and calculates the deadline by combining the effective capacity E and the reputation value R; Step C2. performs proposal broadcast: Firstly traversing and searching a random number meeting the difficulty requirement and taking an edge node reaching the deadline as a winner in the first stage, and broadcasting a Phase1 proposal; Step D2. consensus voting: If the signature votes are legal, the winner must collect the signature votes of more than 2/3 members of the consensus group in the first stage, and the block is successfully verified; step E2, block out: after consensus is reached, the first Phase winner broadcasts a Phase1 acknowledgment block message.
  4. 4. The system of claim 1 wherein the second stage competition out block module competition out block process is to exclude the first stage winner, the remaining K-1 consensus group members use hash values of Phase1 block as new seeds, recalculate challenge values, and perform traversal search in a locally stored capacity map to find new best random numbers matching the challenge values, and calculate deadlines in combination with effective capacity E and reputation value R, then first traverse edge nodes that retrieve random numbers meeting difficulty requirements and reach deadlines as second stage winners, broadcast Phase2 proposals, then the consensus group members verify Phase2 proposals and vote, the second stage winner also has to collect signature votes exceeding 2/3 consensus group members, and finally, after consensus is achieved, the second stage winner broadcasts Phase2 confirmation block messages.
  5. 5. The internet of vehicles traffic information consensus system according to claim 1 wherein the dual-stage differentiated rewarding module gives a fixed base rewarding for the first stage winner with a rewarding value of 2 and a capacity matching dynamic rewarding for the second stage winner with a rewarding formula of: ; Wherein Reward _p2 is the prize value of the second stage winner, e_1 is the effective capacity of the Phase1 winner, e_2 is the effective capacity of the Phase2 winner, and the base prize value BaseReward is 2.
  6. 6. The internet of vehicles traffic information consensus system according to claim 1, wherein the reputation management and self-healing module awards rewards to edge nodes that successfully go out blocks and participate in effective voting, and the reputation value updating process is as follows: ; Wherein Δr obeys a normal distribution N (2.0,0.5) with a mean value of 2.0 and a standard deviation of 0.5, r_old is the reputation value of the last state of the edge node, if the edge node is a successful block node, the value is the reputation value added with the reward value, and r_new is the updated reputation value; Punishment is given to the edge node which is detected to be double-blocked or malicious, and the reputation value updating process is as follows: ; ; wherein Penalty is a penalty value, currentRep is the current reputation value of the edge node, and Threshold is a reputation Threshold.
  7. 7. The traffic information consensus system of the internet of vehicles according to claim 1, wherein a double-layer blacklist mechanism is built in the reputation management and self-healing module, global big-privilege and self-healing logic is introduced, the double-layer blacklist comprises a temporary observation list Level-1 and a permanent forbidden list Level-2, wherein the triggering condition of the temporary observation list Level-1 is that an edge node reputation value is lower than a threshold value for the first time or overtime for a plurality of times, when the Level-1 is triggered, the corresponding edge node is cancelled to participate in consensus group competitive qualification, a recovery time trigger is started, if no further malicious behavior exists in an observation period, the edge node is moved out of the temporary observation list and the corresponding reputation is recovered, the triggering condition of the permanent forbidden list Level-2 is that the edge node is averted or counterfeited again during the Level-1, and when the Level-2 is triggered, the corresponding edge node is permanently deprived of network rights; The global privilege and self-healing logic is that when the system detects that the network throughput is recovered to be normal or the proportion of malicious nodes is reduced to exceed a set value, and meanwhile, the average credit is lower than a set threshold value, a global recovery event is triggered, global privilege protocols are executed by the edge nodes of the whole network, a local edge node list is traversed, credit values of all edge nodes which are in a temporary observation list Level-1 but do not enter a permanent forbidden list Level-2 are forcedly reset to be a safety datum line, and qualification of participating in Top50% consensus group auction is obtained again in the next round.
  8. 8. The internet of vehicles traffic information consensus system according to claim 3 wherein the expression of the effective capacity E in step A2 is: ; The expression of the Deadline loadline in step B2 is: Wherein BaseTime is the base time.
  9. 9. A method of route planning based on a consensus system, the method comprising the steps of: step 1, acquiring dynamic traffic data in real time by an on-board unit OBU, a road side sensor and a pedestrian/non-motor vehicle unit, and transmitting the acquired original road condition data to a nearby road side unit RSU, namely an edge node; Step 2, the road side unit RSU acquires road condition information, submits the road condition information to an authentication layer after identity verification and data preprocessing are completed, the authentication layer carries out consensus based on the vehicle networking road condition information consensus system according to any one of claims 1 to 8, and when the consensus threshold is reached, data uplink is completed, and a determined road condition state snapshot is generated; Step 3, uploading the generated determined road condition state snapshot to an application service layer, and providing a unified and agreed road condition state snapshot for a dynamic path planning module in the application service layer; And 4, carrying out path planning by the dynamic path planning module based on the road condition state snapshot.
  10. 10. The route planning method based on the consensus system according to claim 9, wherein the consensus process of the internet of vehicles road condition information consensus system in step 2 is as follows: Step 2.1, selecting a coordinator of a consensus process according to whether the coordinator is in an initialization stage, starting a mortgage window by broadcasting a message by the coordinator in each round, respectively calculating self comprehensive scores by the edge nodes of the whole network and sending the comprehensive scores to the coordinator, after receiving replies, excluding the edge nodes in a double-layer blacklist by the coordinator, sequencing from high to low according to the scores, intercepting the front 50%, finally selecting the front K from the front 50% of edge nodes by the coordinator according to a logarithmic formula K=6Xlog2 (N), and forming a consensus group and broadcasting a list; step 2.2.Phase1 minimum deadline competition: After receiving the list, the K consensus group members start local PoC minimum deadline competition, each edge node generates a challenge value according to the last block hash, quickly searches the best random number in the local Plot file, calculates the deadline, and if the edge node A finishes searching first and reaches the deadline, packages the transaction to generate a Phase1 proposal block and broadcasts the proposal block to the whole network; Step 2.3.Phase1 validation and phase transition: The rest K-1 consensus group members receive the proposal block of the edge node A, verify the legitimacy, send signature information to the edge node A after the verification is passed, broadcast Phase1 confirmation block after the edge node A collects the signature exceeding 2/3 XK, and automatically enter Phase2 after the whole network edge node receives the confirmation; step 2.4.Phase2 double acknowledgement: The rest K-1 consensus group members start a second round of PoC search based on Phase1 block hash, the edge node C obtains Phase2 block out weight, generates a Phase2 proposal containing edge node A block hash references, re-votes and confirms the consensus group, broadcasts Phase2 final confirmation blocks after consensus is achieved, completes data uplink, and generates a determined road condition state snapshot; Step 2.5, reputation settlement and self-healing: And then updating the reputation of the edge nodes of the whole network based on a dynamic reputation scoring model, synchronously updating a temporary observation list Level-1 and a permanent forbidden list Level-2, calculating an average reputation, and automatically triggering reputation restoration by the system if the system detects that the network throughput is restored to be normal or the proportion of malicious nodes is reduced to exceed a set value and the average reputation of the network is lower than a set threshold value at the moment, and resetting the reputation of all the edge nodes in the temporary observation list to be 70.0 to complete self-healing.

Description

Internet of vehicles road condition information consensus system and route planning method based on consensus system Technical Field The invention belongs to the technical field of internet of vehicles data processing, and particularly relates to an internet of vehicles road condition information consensus system and a route planning method based on the consensus system. Background In typical internet of vehicles applications, such as intersections or specific road segments, data sources from multiple vehicles, multiple sensors, and the like, can generate large amounts of real-time traffic information, even in conflict with each other. For example, there may be multiple reports of different time stamps or location descriptions for the same event (e.g., a temporary obstacle). If the navigation and route planning system makes decisions directly based on these raw data, the problems of frequent jitter of the route, inaccurate estimated arrival time (ETA) and the like will result. Wu Hanlin et al propose a fine-grained path planning method based on Deep Learning (DL), by which traffic space-time characteristics of vehicles are predicted (reference ：DeepLearningEnabledFineGrainedPathPlanningforConnectedVehicularNetworks,IEEETransactionsonVehicularTechnology71.10(2022):10303-10315.）,, although it can realize prediction of future road traffic trend, can not share real-time road condition information, and discounts the timeliness of road conditions greatly. OubbatiOmarSami et al propose a vehicle path planning scheme based on SDN architecture (reference ：SEARCH:anSDN-EnabledApproachforVehiclePath-Planning,IEEETransactionsonVehicularTechnology69.12(2020):14523-14536.）, is used for enhancing situation awareness of urban roads, efficiently collecting traffic information in real time and formulating an optimal navigation strategy, but the vehicle path planning scheme is formed by unmanned aerial vehicles, does not conform to the common condition of daily traffic processes, and cannot solve the data collection problem before general path planning. Chen et al propose an automatic driving vehicle path planning scheme using deep reinforcement learning on network edge nodes (reference ：AnIntelligentPathPlanningSchemeofAutonomousVehiclesPlatoonUsingDeepReinforcementLearningonNetworkEdge,IEEEAccess8(2020):99059-99069.）, can obviously reduce fuel consumption of a vehicle team while guaranteeing task deadline, but an actual object is an automatic driving vehicle, so that the automatic driving technology is still immature, and therefore, the universality of the automatic driving vehicle path planning scheme to actual road conditions is insufficient, and the actual complex road conditions cannot be met. The consensus mechanism can ensure the consistency of the block chain account book data, consensus the real-time road condition data of the Internet of vehicles, provide highly reliable and consistent data for an application decision layer, and solve the problems of frequent jitter of paths, inaccurate estimated arrival time (ETA) and the like. However, the consensus performance requirements in different scenes are different, and a consensus mechanism needs to be designed or selected according to scene characteristics and requirements. The conventional capacity demonstration (PoC) consensus mechanism (e.g., burstcoin, chia) performs minimum time competition by using hard disk space resources instead of power resources, has the remarkable advantage of green energy saving, but has the main disadvantages of 1) high acknowledgement delay, namely that the conventional PoC generally requires a long block-out interval due to the need of preventing bifurcation, resulting in low system Throughput (TPS). 2) Security problems are susceptible to both "no interest" (NothingatStake) attacks and "Grinding" attacks because of the extremely low cost of creating multiple block branches. 3) The existing consensus system is lack of dynamic adaptability, and after being subjected to large-scale network attacks (such as DDoS or Sybil attacks), the existing consensus system is often lack of an automatic recovery mechanism. Even if the attack stops, the honest node may not be able to timely return to the consensus process due to damaged reputation or resource exhaustion, resulting in a system in a low performance state for a long period of time. 4) The expansibility is limited, and the communication overhead of the whole network broadcasting and verification increases exponentially with the increase of the number of the whole network nodes. Therefore, a consensus method suitable for the environment of the internet of vehicles, which not only can keep the PoC energy-saving characteristic, but also can remarkably improve the transaction processing speed, has the anti-attack capability and can self-heal in a large-scale network is needed. Disclosure of Invention In view of the drawbacks and shortcomings of the prior art, a first object of the present invent