CN114500046B - Consensus method based on field network and IOTA - Google Patents

Abstract

The invention discloses a consensus algorithm based on a field network and an IOTA, which comprises the steps of firstly establishing an IOTA consensus time analysis model in the field network; the IOTA consensus time calculation in the field network comprises the time of writing a transaction into a log, the time of transmitting the transaction in the network and the consensus time; then, a parasitic chain attack success probability analysis model is established, and on the basis of the known arrival rate of the honest node event, the ratio of the calculation force required to be paid by the malicious node to the total calculation force of the honest node in the network is analyzed and calculated if the malicious node wants to attack successfully; finally, based on the analysis model, a new block chain consensus algorithm for reducing consensus time by reducing the log writing time is provided, and an IOTA node and adjacent terminals form a cluster; all members of the cluster cooperate to complete tasks and return calculation results to the IOTA node, and then the IOTA node writes data into the list, so that the consensus time of the network is reduced.

Description

Consensus method based on field network and IOTA

Technical Field

The invention relates to the technical field of wireless communication, in particular to a consensus method based on a field network and an IOTA.

Background

With the large-scale development of renewable energy distributed generation, the next generation power grid, i.e., the smart grid, has been developed by integrating the current information network and information system with the conventional energy network into a new power grid system. However, current smart grids still face challenges such as lack of efficient transactional trust between consumers and renewable energy generation, difficulty in achieving dynamic balance of energy supply and demand, emergence of cyber physical security threats, and security of large-scale data sharing. To address these challenges, many researchers are studying the applicability of blockchain technology in SG. For example, there are some important application areas of blockchains in SG, which can be effectively solved: 1) enabling trusted distributed energy transactions, 2) overcoming privacy and security related challenges, 3) remotely controlling energy flow in a particular area by monitoring usage statistics of the particular area, and 4) facilitating diagnosis and maintenance of SG equipment.

However, in SG with millions of distributed wireless terminals, blockchain selection is an important issue. It is well known that coherency algorithms play a key role in the security and coherency of blockchains. Current popular consensus algorithms include workload certification (PoW), rights and interests certification (PoS), and Practical Bayer Fault Tolerance (PBFT). The core idea of PoW is competition for computing power. The miners continue to hash to contend for the right to generate additional blocks, and thus this is not the case for the less computationally intensive SG. The PoS determines the probability of a participant getting accounting rights by evaluating the number and duration of tokens held. In PoS, the higher the coin-age the higher the probability that a participant gets the right to create a new block, which may lead to monopolization. In PBFT, there is one master node in the PBFT consensus mechanism, and the remaining nodes are generally called replicas. The client sends requests to the master node, and then all honest replica nodes execute the order of all transaction requests in the blockchain network by running a protocol consisting of three phases, pre-preparation, preparation and commit. The PBFT consensus algorithm has a faster consensus speed, but needs to know the number of all participants in advance, has difficult capacity expansion and is not suitable for SG.

Therefore, due to the inherent characteristics of large-scale SG networks, fast access, limited computing resources, frequent transactions, etc., the conventional consensus algorithm cannot be effectively applied to the networks. Therefore, some consensus algorithms that are resource efficient, low cost, high transaction throughput, and support concurrency need to be studied. The IOTA is developed specifically for the internet of things, has a Directed Acyclic Graph (DAG) consensus mechanism, breaks the single-chain structure of the consensus mechanism, and allows all nodes to insert new transactions into the distributed ledger immediately after processing and verifying the previous transactions. Tangle is a third generation blockchain based on DAG architecture. The advantages of Tangle are: 1) No transaction fee; 2) Reducing resources; 3) Reducing verification delay of the transaction; 4) Greater network throughput. The IOTA system is thus well suited for SG networks.

However, current research on this problem is mainly focused on specific consensus algorithms, lacking a consensus time analysis model for different network scales and different transmission protocols.

Disclosure of Invention

The invention aims to: aiming at the problems in the background technology, the invention provides a consensus algorithm based on a field network and an IOTA, comprehensively considers the number of wireless transmission hops, the communication protocol and the number of selected tips, establishes a successful probability analysis model of the field network consensus time and parasitic chain attack, solves the problem of secure sharing of SG network data and transaction, and on the basis, provides a new blockchain consensus algorithm, reduces transaction delay and improves system throughput.

The technical scheme is as follows: in order to achieve the above purpose, the invention adopts the following technical scheme:

a consensus method based on a field network and an IOTA comprises the following steps:

step S1, establishing an IOTA consensus time analysis model in a field network; the IOTA consensus time calculation in the field network comprises the time of writing a transaction into a log, the time of transmitting the transaction in the network and the consensus time; in particular, the method comprises the steps of,

s1.1, trading the time of writing a log;

the time of transaction writing is related to the PoW difficulty and the computing power of the device; when a node writes a transaction into a log, firstly selecting the transaction which is not verified in the log as tips for verification, then performing PoW calculation, finally writing data into a local log, and broadcasting the calculation result and the selected tips into an IOTA network; ignoring the verification time of tips, and taking the calculated time of PoW as the time of writing the transaction into the log;

step S1.2, the time of the transaction transmitted in the network;

selecting a Gossip protocol for transaction transmission, and periodically forwarding the transaction owned by the node but not owned by the adjacent node to the adjacent node; repeating the forwarding step by the neighbor node which receives the forwarding message; finally, the transaction information is transmitted to the whole network through the forwarding of a plurality of rounds; the transmission time of the transaction is the transmission hop count of the transaction in the network multiplied by the transmission delay of each hop; the transmission delay of the wired network is ignored, and the transmission delay of each hop of the network in the Field Area Network (FAN) (Field Area Network) is the transmission delay in the mesh of the wireless mesh network; s1.3, consensus achievement time;

The time to reach consensus is the time from when a transaction first exists in all IOTA nodes until each tips can see it, i.e. the time required for tips to be selected for verification; after the IOTA node writes the transaction into the log, when verifying that the tips number of the transaction is equal to the number of the prompts in the network, the consensus is achieved, and the transaction cannot be changed;

s2, establishing a parasitic chain attack success probability analysis model;

the parasitic chain attack success probability analysis model refers to a closed expression of the probability of the success of the parasitic chain attack and the arrival rate of the honest node event and the arrival rate of the malicious node event; in particular, the method comprises the steps of,

as shown in fig. 5, the node firstly transmits a transaction a, and after the transaction is completed consistently, the node receives the materials corresponding to the transaction; at this time, the node sends the double-flower transaction b, when the final accumulated weight of the transaction b is greater than a, the transaction a is abandoned, and the attack is successful at this time; if the parasitic chain selects transaction d initiated after the consensus completion time to verify, transaction b verifies transaction a; each transaction verifying transaction b indirectly verifies transaction a, at which point the parasitic chain attack is unlikely to succeed; the transaction of the parasitic chain selection should therefore be a transaction initiated before the agreed completion time; after the transaction is selected, the attack can be successful only by constructing that the accumulated weight of the parasitic chain is greater than that of the transaction a; calculating the ratio of the calculated amount required by the malicious node to the calculated amount of the whole network when the attack is successful, and obtaining the probability of the attack success;

Step S3, establishing a block chain consensus algorithm based on a field network and an IOTA;

forming a cluster by the IOTA node and the adjacent terminals; the terminal sends the transaction to the IOTA node; when the IOTA node receives enough transactions, it packages all the transactions and distributes the PoW tasks to the terminals in the cluster; and all members in the cluster cooperatively complete respective tasks, the calculation result is returned to the IOTA node, and the IOTA node writes the data into a log.

Further, the specific calculation method of the transaction writing time in step S1.1 is as follows:

setting the node number in the IOTA network as N, when the node i has transaction written into the range, firstly selecting k tips for verification, then performing PoW calculation, finally writing data into the local range, broadcasting the calculation result and the selected tips into the IOTA network, and writing time t of the transaction of the node i _i The method comprises the following steps:

wherein d _Pi To count cycles of PoW, f _i The computing power of node i; the average time interval h of transaction write local log is:

the specific calculation method for the time of the transaction transmitted in the network in the step S1.2 is as follows:

for an IOTA network of N nodes, setting m+1 neighbors of each node, transmitting to m neighbors each time, setting the average time of each hop transmission as T, and setting the probability of successfully receiving an event as p, wherein the network transmission is equivalent to pm fork tree, and the total hop number of transmission is:

L＝log _pm (pmN-N+1)-1

The total transmission time is LT;

based on 802.11MAC broadcast protocol, gaussian channel, when data saturation, calculating average transmission time T of each hop;

when the node has a data packet to be sent, the node starts to monitor the channel state, when the idle time of the channel is detected to exceed the distributed frame interval DIFS, the node enters a back-off stage, otherwise, the node is always in the channel state; after the node enters the backoff stage, randomly selecting a backoff time slot value and starting a backoff counter; when the channel is in an idle state, the back-off counter is decremented along with the idle time of the channel by taking time slots as a unit, and a data packet starts to be sent until the back-off counter is zero; when the channel is busy in the decrementing process of the backoff counter, the backoff counting process is frozen, the node monitors the channel again until the idle time of the channel is detected to exceed DIFS, and the node returns to the backoff stage again and continues the frozen backoff decrementing process;

setting the back-off time setting range to {1-w ₀ Probability τ=2/(w) of node transmitting data ₀ +1), probability p of successful transmission of data in one virtual slot _s ＝nτ(1-τ) ^n-1 The probability of channel idle is p _i ＝(1-τ) ⁿ The probability of channel busy is p _bsy ＝1-p _i ＝1-(1-τ) ⁿ Wherein n is the number of nodes which can generate collision; in a virtual time slot, the probability of collision is p _c ＝1-p _s -p _i When the channel successfully transmits data or collides, the transmission delay of the data comprises: data transmission delay L _en R, wait time delay DIFS and propagation time delay delta before transmission of the distributed frame; wherein L is _en For the total length of the frame, including a MAC data packet header and a physical data packet header, r is the sending rate; time delay T of busy channel due to successful transmission _s And time delay T of busy channel caused by collision _c The method comprises the following steps:

the latency required for each technique of the back-off counter includes: when the channel is idle, the back-off counter is frozen every 1 time, the time interval is one physical time slot sigma, when the channel successfully transmits data, the back-off counter is frozen, and the waiting time is T _s And when the channel collides, the back-off counter is frozen with a waiting time of T _c The method comprises the steps of carrying out a first treatment on the surface of the The back-off counter thus waits for an average of time intervals per timer:

t ₀ ＝(1-p _bsy )σ+p _s T _s +p _c T _c

t ₀ namely a virtual time slot; the back-off timer initial value D is {1-w ₀ All of the } rangesEvenly distributed, E [ D ]]＝(w ₀ +1)/2; for n nodes that may collide, the timer performs 1-down counting simultaneously after setting the initial value, and the back-off time E [ D ] divided equally to each node _n ]＝(w ₀ +1)/2 n, the average delay for transmitting a data frame is:

setting the data packet size of the IOTA as L _p One data frame has a size L _en The average time to transmit an IOTA packet is:

calculating the probability of successful reception of the IOTA data packet by adopting M-ary QAM coding; the bit error rate is expressed as follows:

wherein E is _b /N ₀ For the signal-to-noise ratio,

the probability of error by data transmission is p _e ＝1-(1-p _b ) ^Len The data transmission failure includes a data transmission error or collision, and therefore, the probability of the data transmission failure is: p is p _r ＝p _c +p _e -p _c *p _e The probability of successful transmission of an IOTA packet is:

the consensus achievement time calculation method in the step S1.3 is as follows:

in the range, the total of tipsNumber R ₀ Is maintained in a stable state and satisfies:

at the time t-h, tips number is R ₀ At time t, the number of tips is still R ₀ And in the h period, the event arrival number is lambdah, whereinThus, for a transaction that is tips at time t-h, the probability of having been verified at time t is +.>

Setting at a time t-h, and verifying that the tips number of the sites i is phi (t-h); wherein the transactions that have been validated are referred to as sites i, and the transactions that have not been validated are referred to as tips i; tips at time t-h, and the verified set at time t is: |a|=pψ (t-h) = (k-1) ψ (t-h)/k; tips at time t-h, and the set of tips still at time t is: i B i= (1-p) ψ (t-h) =ψ (t-h)/k; the tips set of sites i is not verified to be |C| at the time t-h; at the time t-h, when a transaction arrives, k tips are selected from all tips to verify, if the selected tips are in the psi (t-h), the newly arrived transaction indirectly verifies site i; if k tips are all selected from the set B, the number of tips of the verification sites i is reduced by k-1; k tips are selected from the set B to be k-1, one of A or C is selected, the number of tips of the verification sites i is reduced by k-2, and the method is similar, none of the k tips is selected from the set B, and at least one of the k tips is selected from the set A, and then the number of tips of the verification sites i is increased by 1 at the moment t; the following probabilities can be found:

+1：

-1：

-2：

……

-(k-1)：

NeglectingTo obtain the growth rate of ψ (t):

since ψ (0) =1, the settings are setIs available in the form of

Wherein,is a Langmuir omega function;

namely:

and (3) making:

when ψ (t) =r ₀ When the site i reaches consensus, the consensus reaching time is:

when the network transmission delay is ignored, the consensus time is as follows:

when the network transmission time delay is added, at a certain moment, a transaction is written into a log by a node i in the network, and before the tips i are transmitted to other nodes, the tips i are only visible in the local log of the node i, and when other new transactions arrive, only the transaction arriving in the node i can select the tips i as a father node; when data is put into a local list but is not transmitted to a neighbor node, namely T < T, the number of tips of the verification tips i is psi (T) = (1/N) exp (at/h), when T <2T, the number of tips of the verification tips i is psi (T) = (1/N) exp (at/h) + (pm/N) exp (a (T-T)/h), and the like, the number of tips of the verification tips i is available in the period from the first writing of the list to the completion of broadcasting in a network, wherein the number of tips of the verification tips i is:

order the

After the broadcasting is completed: the value of psi (t) at the time of LT is the initial value of t.gtoreq.LT,

wherein t is ₀ For the time elapsed from the data write to completion of consensus, when ψ (t) =r ₀ At the time, site i achieves consensus, so consensus achieves time t _c The method comprises the following steps:

under the condition that the network transmission delay is not negligible, the consensus time is:

further, the specific method for establishing the parasitic chain attack success probability analysis model in the step S2 is as follows:

setting the event arrival rate of honest node as lambda and the event arrival rate of malicious node as mu, and setting the cumulative weight as phi (t) when the transaction a reaches consensus ₀ ) Wherein t is ₀ The time it takes for a transaction to write a log to arrive at a consensus;

for honest nodes, after consensus is achieved, as things issued by each honest node will verify tips which are already consensus, the cumulative weight growth rate is lambda and the cumulative weight growth rate of transaction b is mu after consensus is achieved for transaction a; setting the release time of the transaction a as T ₁ Parasitic chain start time T ₂ The method comprises the steps of carrying out a first treatment on the surface of the If T ₂ After the transaction a has agreed to completion time t, then transaction b will verify transaction a, i.e. verify transaction bEach transaction will indirectly verify transaction a, and parasitic link attacks are unlikely to succeed, so the parasitic link start time should be less than the consensus completion time t ₀ The method comprises the steps of carrying out a first treatment on the surface of the When the trade a consensus is reached, the cumulative weight of the parasitic chain is μ (t ₀ -T ₂ ) The cumulative weight of honest nodes is phi (t ₀ -T ₁ ) The method comprises the steps of carrying out a first treatment on the surface of the Thus, if the attack is successful, after consensus is reached, the parasitic chain cumulative weight should be greater than that of the honest node, i.e., φ (t ₀ -T ₁ )-μ(t ₀ -T ₂ )<0; setting T ₁ ＝T ₂ =0, i.e. the parasitic chain starts to be ready when transaction a writes to log, then phi (t ₀ )-μ(t ₀ )<0; the cumulative weights phi (t) at the completion of consensus are calculated respectively ₀ ) The preparation is started at the moment of writing a log with the parasitic chain transaction a, and the probability of attack success is increased;

cumulative weight at transaction consensus completion: when a new transaction arrives, at least one verification site i exists in the k selected tips, and the new transaction indirectly verifies the site i, wherein the accumulated weight of the site i is added with 1; at least one tips verifies that the probability of site i is 1- (1-psi (t-h)/R ₀ ) ^k

Wherein phi (t) is the cumulative weight;

the method can obtain:

wherein,

by the nature of the omega functionThe method can obtain:

thus, the cumulative weight φ (t) can be calculated as:

since the weight accumulation time is mainly in the data transmission stage and the consensus achievement stage, t in the above formula is the sum of the data transmission time and the accumulated weight time, t ₀ The cumulative weight is expressed as follows:

probability of success of parasitic chain attack: because the honest nodes accord with poisson distribution with the event arrival rate lambda, the time intervals of adjacent things of the honest nodes are index variables which are independently and uniformly distributed and obey 1/lambda index distribution; similarly, the time interval between adjacent things of the malicious node is 1/mu of independent index variable distributed in the same way; the method can obtain:

When mu>Lambda, when mu is less than or equal to lambda, only when the transaction a agrees, the accumulated weight of the transaction b is greater than that of the transaction a, so that the attack is possible to be successful, namely phi (t ₀ )-μ(t ₀ )>0; upon successful attack, the malicious node should track the difference φ (t) of the cumulative weights of the parasitic chain and the validation transaction a ₀ )-μ(t ₀ ) Therefore, the probability of success of a parasitic chain attack is:

further, the establishing a blockchain consensus algorithm based on the field network and the IOTA in the step S3 specifically includes: an IOTA node and adjacent terminals are formed into a cluster. The terminal sends a transaction to the IOTAnode. When the IOTA node receives enough transactions, it will package all the transactions and assign PoW tasks to the terminals in the cluster. All members of the cluster cooperate to complete their respective tasks and return the results to the iotaonode, which then writes the data to a chaotic node. Thus, the amount of IOTAnode effort, as well as the time required to write data to chaotic data, can be significantly reduced.

After clustering, after waiting for the transaction to reach the maximum value of the IOTA packet, the data can be written into the log, and the data collection time is as follows:

wherein n is _t N is the average transaction number in one data packet _g Lambda is the number of nodes in a cluster _i Event arrival rate for node i; the time of writing data into the table is mainly the PoW calculation time, and after clustering, poW is completed cooperatively by members in the cluster, so that the time of writing data into the table is as follows:

wherein d _Pi For counting cycles of PoW calculation, F _i Computing power for cluster iThe average time interval for the event to write to the local log is:

because the data only needs to be transmitted to the IOTAnode node in the network, the transmission time of the data in the network is less than or equal to the transmission time when the data is not clustered, and the maximum transmission time LT is taken here;

the consensus achievement time is:

wherein->

The consensus time is expressed as follows:

the beneficial effects are that:

the consensus algorithm based on the field network and the IOTA provided by the invention establishes an IOTA analysis model aiming at the large-scale distributed wireless network of the SG under the condition of considering network transmission delay, provides a basis for block chain application analysis of the distributed wireless network, further derives a closed expression of consensus time of parasitic chain attack success probability, provides a basis for calculation of block chain consensus time of the distributed wireless network and probability analysis of parasitic chain attack, and finally provides a new consensus algorithm for completing PoW by node clustering and cooperation among nodes, thereby reducing IOTA consensus time and improving transaction efficiency.

Drawings

FIG. 1 is a schematic view of a Tangle structure;

FIG. 2 is a system model diagram of a consensus method provided by the present invention;

FIG. 3 is a schematic diagram of a tips selection algorithm in an embodiment of the invention;

FIG. 4 is a schematic diagram of tips locations selected by a parasitic chain attack in an embodiment of the present invention;

fig. 5 is a schematic diagram of a parasitic link attack time point in an embodiment of the present invention.

Detailed Description

The invention will be further described with reference to the accompanying drawings. It will be apparent that the described embodiments are some, but not all, embodiments of the invention. All other embodiments, which can be made by those skilled in the art based on the embodiments of the invention without making any inventive effort, are intended to be within the scope of the invention.

The invention comprehensively considers the hop count, the communication protocol and the option tips number of the data wireless transmission, establishes a successful probability analysis model of the field network consensus time and the parasitic chain attack, and provides a new block chain consensus method based on the successful probability analysis model, as shown in figure 2. The method specifically comprises the following steps:

and S1, establishing an IOTA consensus time analysis model in the field network. The IOTA consensus time calculation in the field network comprises the time of writing a transaction into a log, the time of transmitting the transaction in the network and the consensus time. In particular, the method comprises the steps of,

Step S1.1, trading the time of writing the log.

The time the transaction writes to the log correlates with the amount of work done to prove PoW difficulty and the computing power of the device. When the node has transaction written in the list, firstly, tips are selected for verification, then PoW calculation is carried out, finally, data are written in the local list, and the calculation result and the selected tips are broadcasted to the IOTA network. Since the verification time of tips is negligible compared with the calculation time of PoW, the time required from the selection of tips from node i to the smooth writing of events into local tangles is mainly the calculation time of PoW, and the calculation time of PoW is taken as the time of writing transactions into tangles.

Setting the node number in the IOTA network as N, when the node i has transaction written into the range, firstly selecting k tips for verification, then performing PoW calculation, finally writing data into the local range, broadcasting the calculation result and the selected tips into the IOTA network, and writing time t of the transaction of the node i _i The method comprises the following steps:

wherein d _Pi To count cycles of PoW, f _i Is the computational power of node i. The average time interval h of transaction write local log is:

step S1.2, the time of the transaction transmitted in the network.

And selecting a Gossip protocol for transaction transmission, and forwarding the transaction owned by the node and not owned by the adjacent node to the adjacent node periodically. The neighbor node that received the forwarding message repeats the forwarding step. And finally, the transaction information is transmitted to the whole network through the forwarding of a plurality of rounds. The transmission time of a transaction is the number of transmission hops of the transaction in the network multiplied by the transmission delay per hop. The transmission delay of the wired network is ignored, and the transmission delay of each hop of the network in the FAN is the transmission delay in the mesh.

For an IOTA network of N nodes, setting m+1 neighbors of each node, transmitting to m neighbors each time, setting the average time of each hop transmission as T, and setting the probability of successfully receiving an event as p, wherein the network transmission is equivalent to pm fork tree, and the total hop number of transmission is:

L＝log _pm (pmN-N+1)-1

the total transmission time is LT.

Based on the 802.11MAC broadcast protocol, the gaussian channel, at the time of data saturation, the average transmission time T per hop is calculated.

When the node has data packets to send, the node starts to monitor the channel state, when the idle time of the channel is detected to exceed the distributed frame interval DIFS, the node enters a back-off stage, otherwise, the node is always in the monitor channel state. After the node enters the backoff stage, a backoff slot value is randomly selected and a backoff counter is started. When the channel is in an idle state, the back-off counter is decremented by time slot units along with the idle time of the channel until the back-off counter is zero, and the data packet starts to be sent. When the channel is busy during the backoff counter decrementing, the backoff counting process is frozen, the node listens to the channel again until after detecting that the channel idle time exceeds DIFS, reverting back to the backoff stage and continuing the frozen backoff decrementing process.

Setting the back-off time setting range to {1-w ₀ Probability τ=2/(w) of node transmitting data ₀ +1), probability p of successful transmission of data in one virtual slot _s ＝nτ(1-τ) ^n-1 The probability of channel idle is p _i ＝(1-τ) ⁿ The probability of channel busy is p _bsy ＝1-p _i ＝1-(1-τ) ⁿ Where n is the number of nodes that may cause a collision. In a virtual time slot, the probability of collision is p _c ＝1-p _s -p _i When the channel successfully transmits data or collides, the transmission delay of the data comprises: data transmission delay L _en R, the pre-transmission wait time delay DIFS and the propagation time delay delta. Wherein L is _en For the total length of the frame, including the MAC packet header and the physical packet header, r is the transmission rate. Time delay T of busy channel due to successful transmission _s And time delay T of busy channel caused by collision _c The method comprises the following steps:

the latency required for each technique of the back-off counter includes: when the channel is idle, the back-off counter is reduced by 1, the time interval is one physical time slot sigma, and when the channel is successfully transmittedAccordingly, the back-off counter is frozen, and the waiting time is T _s And when the channel collides, the back-off counter is frozen with a waiting time of T _c . The back-off counter thus waits for an average of time intervals per timer:

t ₀ ＝(1-p _bsy )σ+p _s T _s +p _c T _c

t ₀ I.e. a virtual time slot. The back-off timer initial value D is {1-w ₀ Uniformly distributed in the range E [ D ]]＝(w ₀ +1)/2. For n nodes that may collide, the timer performs 1-down counting simultaneously after setting the initial value, and the back-off time E [ D ] divided equally to each node _n ]＝(w ₀ +1)/2 n, the average delay for transmitting a data frame is:

setting the data packet size of the IOTA as L _p One data frame has a size L _en The average time to transmit an IOTA packet is:

and calculating the probability of successful reception of the IOTA data packet by adopting M-ary QAM coding. The bit error rate is expressed as follows:

wherein E is _b /N ₀ For the signal-to-noise ratio,

the probability of error of data transmission isThe data transmission failure includes a data transmission error or collision, and therefore, the probability of the data transmission failure is: p is p _r ＝p _c +p _e -p _c *p _e The probability of successful transmission of an IOTA packet is:

step S1.3, consensus achievement time.

The time to reach consensus is the time from when a transaction first exists in all IOTA nodes until each tips can see it, i.e., the time required for tips to be selected for verification. After the IOTA node writes the transaction into the log, when the number of tips of the transaction is verified to be equal to the number of the prompts in the network, the consensus is achieved, and the transaction cannot be changed.

In the range, the total number of tips R ₀ Is maintained in a stable state and satisfies:

at the time t-h, tips number is R ₀ At time t, the number of tips is still R ₀ And in the h period, the event arrival number is lambdah, whereinThus, for a transaction that is tips at time t-h, the probability of having been verified at time t is +.>

As shown in FIG. 3, the number of tips of the verification sites i is set to be ψ (t-h) at the time t-h. Where the transactions that have been validated are referred to as sites i and the transactions that have not been validated are denoted as tips i. Tips at time t-h, and the verified set at time t is: |a|=pψ (t-h) = (k-1) ψ (t-h)/k. Tips at time t-h, and the set of tips still at time t is: i B i= (1-p) ψ (t-h) =ψ (t-h)/k. At the time t-h, not verifying that tips set of sites i is |C|; at time t-h, when a transaction arrives, k tips are selected from all tips to verify, if the selected tips are in psi (t-h), the newly arrived transaction indirectly verifies site i. If k tips are all selected from set B, the number of tips of verification sites i is reduced by k-1. k tips are selected from the set B to be k-1, one is selected from A or C, the number of tips of the verification sites i is reduced by k-2, and the like, none of the k tips is selected from B, and at least one of the k tips is selected from A, so that the number of tips of the verification sites i is increased by 1 at the moment t. The following probabilities can be found:

+1：

-1：

-2：

……

-(k-1)：/>

NeglectingTo obtain the growth rate of ψ (t):

since ψ (0) =1, the settings are setIs available in the form of

Wherein,as a langerhan function.

Namely:

and (3) making:

when ψ (t) =r ₀ When the site i reaches consensus, the consensus reaching time is:

when the network transmission delay is ignored, the consensus time is as follows:

when the network transmission time delay is added, at a certain moment, a transaction is written into a log by a node i in the network, and before the tips i are transmitted to other nodes, the tips i are only visible in the local log of the node i, and when other new transactions arrive, only the transaction arriving in the node i can select the tips i as a father node. When data is put into a local list but is not transmitted to a neighbor node, namely T < T, the number of tips of the verification tips i is psi (T) = (1/N) exp (at/h), when T <2T, the number of tips of the verification tips i is psi (T) = (1/N) exp (at/h) + (pm/N) exp (a (T-T)/h), and the like, the number of tips of the verification tips i is available in the period from the first writing of the list to the completion of broadcasting in a network, wherein the number of tips of the verification tips i is:

/>

order the

After the broadcasting is completed: the value of psi (t) at the time of LT is the initial value of t.gtoreq.LT,

wherein t is ₀ For the time elapsed from the data write to completion of consensus, when ψ (t) =r ₀ At the time, site i achieves consensus, so consensus achieves time t _c The method comprises the following steps:

under the condition that the network transmission delay is not negligible, the consensus time is:

and S2, establishing a parasitic chain attack success probability analysis model.

The parasitic link attack success probability analysis model refers to a closed expression of the probability of the success of the parasitic link attack and the arrival rate of the honest node event and the arrival rate of the malicious node event. In particular, the method comprises the steps of,

as shown in fig. 5, the node first sends a transaction a, and after the transaction is completed consistently, the node receives the material corresponding to the transaction. At this time, the node sends a double-flower transaction b, and when the final accumulated weight of the transaction b is greater than a, the transaction a is abandoned, and the attack is successful. If the parasitic chain selects transaction d initiated after the consensus completion time to verify, transaction b verifies transaction a. Each transaction that verifies transaction b indirectly verifies transaction a, at which point the parasitic chain attack is unlikely to succeed. The transaction of the parasitic chain selection should therefore be a transaction initiated before the agreed upon completion time. After the transaction is selected, the attack can be successful only by constructing that the accumulated weight of the parasitic chain is larger than that of the transaction a. And calculating the ratio of the calculated amount required by the malicious node to the calculated amount of the whole network when the attack is successful, and obtaining the probability of the attack success.

Setting the event arrival rate of honest node as lambda and the event arrival rate of malicious node as mu, and setting the cumulative weight as phi (t) when the transaction a reaches consensus ₀ ) Wherein t is ₀ The time it takes for a transaction to write a log to arrive in a consensus.

For honest nodes, after consensus is achieved, since everything issued by each honest node will verify the consensus tips, the cumulative weight growth rate for transaction a is λ and the cumulative weight growth rate for transaction b is μ after consensus is achieved. As shown in FIG. 4, the time for issuing transaction a is set to T ₁ Parasitic chain start time T ₂ . If T ₂ After the consensus completion time t for transaction a, then transaction b will verify transaction a, i.e., each transaction verifying transaction b will indirectly verify transaction a, and parasitic link attacks are unlikely to succeed, so the parasitic link start time should be less than the consensus completion time t ₀ . Accumulation of parasitic chains upon achievement of transaction a consensusThe weight is mu (t) ₀ -T ₂ ) The cumulative weight of honest nodes is phi (t ₀ -T ₁ ). Thus, if the attack is successful, after consensus is reached, the parasitic chain cumulative weight should be greater than that of the honest node, i.e., φ (t ₀ -T ₁ )-μ(t ₀ -T ₂ )<0. Setting T ₁ ＝T ₂ =0, i.e. the parasitic chain starts to be ready when transaction a writes to log, then phi (t ₀ )-μ(t ₀ )<0. The cumulative weights phi (t) at the completion of consensus are calculated respectively ₀ ) And the parasitic chain transaction a starts to prepare at the moment of writing the log, and the probability of attack success is increased.

Cumulative weight at transaction consensus completion: when a new transaction arrives, at least one verification site i in the k selected tips is verified, and the new transaction indirectly verifies the site i, and the accumulated weight of the site i is added by 1. At least one tips verifies that the probability of site i is 1- (1-psi (t-h)/R ₀ ) ^k

Where φ (t) is the cumulative weight.

The method can obtain:

wherein,

by the nature of the omega functionThe method can obtain:

thus, the cumulative weight φ (t) can be calculated as:

since the weight accumulation time is mainly in the data transmission stage and the consensus achievement stage, t in the above formula is the sum of the data transmission time and the accumulated weight time, t ₀ The cumulative weight is expressed as follows:

probability of success of parasitic chain attack: because the honest nodes accord with poisson distribution with the event arrival rate lambda, the time intervals of adjacent things of the honest nodes are index variables which are independently and uniformly distributed and obey 1/lambda index distribution. Similarly, the interval between adjacent malicious nodes is 1/mu of independent index variable distributed in the same way. The method can obtain:

when mu>Lambda, when mu is less than or equal to lambda, only when the transaction a agrees, the accumulated weight of the transaction b is greater than that of the transaction The cumulative weight of easy a is the probability that the attack will succeed, i.e., phi (t ₀ )-μ(t ₀ )>0. Upon successful attack, the malicious node should track the difference φ (t) of the cumulative weights of the parasitic chain and the validation transaction a ₀ )-μ(t ₀ ) Therefore, the probability of success of a parasitic chain attack is:

and step S3, establishing a block chain consensus algorithm based on the field network and the IOTA.

And forming the cluster by the IOTA node and the adjacent terminals. The terminal sends the transaction to the IOTA node. When the IOTA node receives enough transactions, it packages all the transactions and distributes the PoW tasks to the terminals in the cluster. And all members in the cluster cooperatively complete respective tasks, the calculation result is returned to the IOTA node, and the IOTA node writes the data into a log.

The algorithm mainly reduces the consensus time by reducing the log writing time, and specifically comprises the following steps: the consensus time is reduced by reducing the tan write time. The method comprises the steps of clustering nodes in a network, wherein each cluster comprises one IOTA node and a plurality of terminals, the terminals in the cluster send data to the IOTA node, when the IOTA node receives enough data, the terminals in the cluster pack all the data, distribute a PoW task to the terminals in the cluster, cooperatively complete the PoW task by all members in the cluster, return a calculation result to the IOTA node, and complete the work of writing data into a column by the IOTA node. Thus, the workload of the IOTA node can be greatly reduced, and the time for writing data into the column is reduced.

After clustering, after waiting for the transaction to reach the maximum value of the IOTA packet, the data can be written into the log, and the data collection time is as follows:

wherein n is _t N is the average transaction number in one data packet _g Lambda is the number of nodes in a cluster _i Is the event arrival rate of node i. The time of writing data into the table is mainly the PoW calculation time, and after clustering, poW is completed cooperatively by members in the cluster, so that the time of writing data into the table is as follows:

wherein d _Pi For counting cycles of PoW calculation, F _i Computing power for cluster iThe average time interval for the event to write to the local log is:

since data only needs to be transmitted to the IOTAnode in the network, the transmission time of the data in the network is less than or equal to the transmission time when the data is not clustered, and the maximum transmission time LT is taken here.

The consensus achievement time is:

wherein->

The consensus time is expressed as follows:

the foregoing is only a preferred embodiment of the invention, it being noted that: it will be apparent to those skilled in the art that various modifications and adaptations can be made without departing from the principles of the present invention, and such modifications and adaptations are intended to be comprehended within the scope of the invention.

Claims (4)

1. The consensus method based on the field network and the IOTA is characterized by comprising the following steps:

step S1, establishing an IOTA consensus time analysis model in a field network; the IOTA consensus time calculation in the field network comprises the time of writing the transaction into the IOTA database, the time of transmitting the transaction in the network and the consensus time; in particular, the method comprises the steps of,

s1.1, trading the time of writing a log;

the time of transaction writing is related to the PoW difficulty and the computing power of the device; when a node writes a transaction into a log, firstly selecting the transaction which is not verified in the log as tips for verification, then performing PoW calculation, finally writing data into a local log, and broadcasting the calculation result and the selected tips into an IOTA network; ignoring the verification time of tips, and taking the calculated time of PoW as the time of writing the transaction into the log;

step S1.2, the time of the transaction transmitted in the network;

selecting a Gossip protocol for transaction transmission, and periodically forwarding the transaction owned by the node but not owned by the adjacent node to the adjacent node; repeating the forwarding step by the neighbor node which receives the forwarding message; finally, the transaction information is transmitted to the whole network through the forwarding of a plurality of rounds; the transmission time of the transaction is the transmission hop count of the transaction in the network multiplied by the transmission delay of each hop; the transmission delay of the wired network is ignored, and the transmission delay of each hop of the network in the field area network FAN is the transmission delay in the wireless mesh network mesh;

S1.3, consensus achievement time;

the time to reach consensus is the time from when a transaction first exists in all IOTA nodes until each tips can see it, i.e. the time required for tips to be selected for verification; after the IOTA node writes the transaction into the log, when verifying that the tips number of the transaction is equal to the number of the prompts in the network, the consensus is achieved, and the transaction cannot be changed;

s2, establishing a parasitic chain attack success probability analysis model;

the parasitic chain attack success probability analysis model refers to a closed expression of the probability of the success of the parasitic chain attack and the arrival rate of the honest node event and the arrival rate of the malicious node event; in particular, the method comprises the steps of,

the node firstly transmits a transaction a, and after the transaction is completed consistently, the node receives materials corresponding to the transaction; at this time, the node sends the double-flower transaction b, when the final accumulated weight of the transaction b is greater than a, the transaction a is abandoned, and the attack is successful at this time; if the parasitic chain selects transaction d initiated after the consensus completion time to verify, transaction b verifies transaction a; each transaction verifying transaction b indirectly verifies transaction a, at which point the parasitic chain attack is unlikely to succeed; the transaction of the parasitic chain selection should therefore be a transaction initiated before the agreed completion time; after the transaction is selected, the attack can be successful only by constructing that the accumulated weight of the parasitic chain is greater than that of the transaction a; calculating the ratio of the calculated amount required by the malicious node to the calculated amount of the whole network when the attack is successful, and obtaining the probability of the attack success;

Step S3, establishing a block chain consensus algorithm based on a field network and an IOTA;

forming a cluster by the IOTA node and the adjacent terminals; the terminal sends the transaction to the IOTA node; when the IOTA node receives enough transactions, it packages all the transactions and distributes the PoW tasks to the terminals in the cluster; and all members in the cluster cooperatively complete respective tasks, the calculation result is returned to the IOTA node, and the IOTA node writes the data into a log.

2. The consensus method based on the field network and the IOTA according to claim 1, wherein the specific calculation method of the transaction writing time in step S1.1 is as follows:

setting the node number in the IOTA network as N, when the node i has transaction written into the range, firstly selecting k tips for verification, then performing PoW calculation, finally writing data into the local range, broadcasting the calculation result and the selected tips into the IOTA network, and writing time t of the transaction of the node i _i The method comprises the following steps:

wherein d _Pi To count cycles of PoW, f _i The computing power of node i; the average time interval h of transaction write local log is:

the specific calculation method for the time of the transaction transmitted in the network in the step S1.2 is as follows:

For an IOTA network of N nodes, setting m+1 neighbors of each node, transmitting to m neighbors each time, setting the average time of each hop transmission as T, and setting the probability of successfully receiving an event as p, wherein the network transmission is equivalent to pm fork tree, and the total hop number of transmission is:

L＝log _pm (pmN-N+1)-1

the total transmission time is LT;

based on 802.11MAC broadcast protocol, gaussian channel, when data saturation, calculating average transmission time T of each hop;

when the node has a data packet to be sent, the node starts to monitor the channel state, when the idle time of the channel is detected to exceed the distributed frame interval DIFS, the node enters a back-off stage, otherwise, the node is always in the channel state; after the node enters the backoff stage, randomly selecting a backoff time slot value and starting a backoff counter; when the channel is in an idle state, the back-off counter is decremented along with the idle time of the channel by taking time slots as a unit, and a data packet starts to be sent until the back-off counter is zero; when the channel is busy in the decrementing process of the backoff counter, the backoff counting process is frozen, the node monitors the channel again until the idle time of the channel is detected to exceed DIFS, and the node returns to the backoff stage again and continues the frozen backoff decrementing process;

Setting the back-off time setting range to {1-w ₀ Probability τ=2/(w) of node transmitting data ₀ +1) in a virtual time slot, the probability of successful transmission of dataRate p _s ＝nτ(1-τ) ^n-1 The probability of channel idle is p _i ＝(1-τ) ⁿ The probability of channel busy is p _bsy ＝1-p _i ＝1-(1-τ) ⁿ Wherein n is the number of nodes which can generate collision; in a virtual time slot, the probability of collision is p _c ＝1-p _s -p _i When the channel successfully transmits data or collides, the transmission delay of the data comprises: data transmission delay L _en R, wait time delay DIFS and propagation time delay delta before transmission of the distributed frame; wherein L is _en For the total length of the frame, including a MAC data packet header and a physical data packet header, r is the sending rate; time delay T of busy channel due to successful transmission _s And time delay T of busy channel caused by collision _c The method comprises the following steps:

the latency required for each technique of the back-off counter includes: when the channel is idle, the back-off counter is frozen every 1 time, the time interval is one physical time slot sigma, when the channel successfully transmits data, the back-off counter is frozen, and the waiting time is T _s And when the channel collides, the back-off counter is frozen with a waiting time of T _c The method comprises the steps of carrying out a first treatment on the surface of the The back-off counter thus waits for an average of time intervals per timer:

t ₀ ＝(1-p _bsy )σ+p _s T _s +p _c T _c

t ₀ Namely a virtual time slot; the back-off timer initial value D is {1-w ₀ Uniformly distributed in the range E [ D ]]＝(w ₀ +1)/2; for n nodes that may collide, the timer performs 1-down counting simultaneously after setting the initial value, and the back-off time E [ D ] divided equally to each node _n ]＝(w ₀ +1)/2 n, the average delay for transmitting a data frame is:

setting the data packet size of the IOTA as L _p One data frame has a size L _en The average time to transmit an IOTA packet is:

calculating the probability of successful reception of the IOTA data packet by adopting M-ary QAM coding; the bit error rate is expressed as follows:

wherein E is _b /N ₀ For the signal-to-noise ratio,

the probability of error of data transmission isThe data transmission failure includes a data transmission error or collision, and therefore, the probability of the data transmission failure is: p is p _r ＝p _c +p _e -p _c *p _e The probability of successful transmission of an IOTA packet is:

the consensus achievement time calculation method in the step S1.3 is as follows:

in the range, the total number of tips R ₀ Is maintained in a stable state and satisfies:

at the time t-h, tips number is R ₀ At time t, the number of tips is still R ₀ And in the h period, the event arrival number is lambdah, whereinThus, for a transaction that is tips at time t-h, the probability of having been verified at time t is +. >

Setting at a time t-h, and verifying that the tips number of the sites i is phi (t-h); wherein the transactions that have been validated are referred to as sites i, and the transactions that have not been validated are referred to as tips i; tips at time t-h, and the verified set at time t is: |a|=pψ (t-h) = (k-1) ψ (t-h)/k; tips at time t-h, and the set of tips still at time t is: i B i= (1-p) ψ (t-h) =ψ (t-h)/k; tips of sites i were not verified to be |C|; at the time t-h, when a transaction arrives, k tips are selected from all tips to verify, if the selected tips are in the psi (t-h), the newly arrived transaction indirectly verifies site i; if k tips are all selected from the set B, the number of tips of the verification sites i is reduced by k-1; k tips are selected from the set B to be k-1, one of A or C is selected, the number of tips of the verification sites i is reduced by k-2, and the method is similar, none of the k tips is selected from the set B, and at least one of the k tips is selected from the set A, and then the number of tips of the verification sites i is increased by 1 at the moment t; the following probabilities can be found:

+1：

-1：

-2：

……

-(k-1)：

neglectingTo obtain the growth rate of ψ (t):

since ψ (0) =1, the settings are setIs available in the form of

Wherein,is a Langmuir omega function;

namely:

and (3) making:

when ψ (t) =r ₀ When the site i reaches consensus, the consensus reaching time is:

When the network transmission delay is ignored, the consensus time is as follows:

when the network transmission time delay is added, at a certain moment, a transaction is written into a log by a node i in the network, and before the tips i are transmitted to other nodes, the tips i are only visible in the local log of the node i, and when other new transactions arrive, only the transaction arriving in the node i can select the tips i as a father node; when data is put into a local list but is not transmitted to a neighbor node, namely T < T, the number of tips of the verification tips i is psi (T) = (1/N) exp (at/h), when T <2T, the number of tips of the verification tips i is psi (T) = (1/N) exp (at/h) + (pm/N) exp (a (T-T)/h), and the like, the number of tips of the verification tips i is available in the period from the first writing of the list to the completion of broadcasting in a network, wherein the number of tips of the verification tips i is:

order the

After the broadcasting is completed: the value of psi (t) at the time of LT is the initial value of t.gtoreq.LT,

wherein t is ₀ For the time elapsed from the data write to completion of consensus, when ψ (t) =r ₀ At the time, site i achieves consensus, so consensus achieves time t _c The method comprises the following steps:

under the condition that the network transmission delay is not negligible, the consensus time is:

3. the consensus method based on the field network and the IOTA according to claim 1, wherein the specific method for establishing the parasitic chain attack success probability analysis model in the step S2 is as follows:

Setting the event arrival rate of honest node as lambda and the event arrival rate of malicious node as mu, and setting the cumulative weight as phi (t) when the transaction a reaches consensus ₀ ) Wherein t is ₀ The time it takes for a transaction to write a log to arrive at a consensus;

for honest nodes, after consensus is achieved, since matters issued by each honest node will verify the consensus tips, the cumulative weight of transaction a is increased at a rate of lambda and the cumulative weight of transaction b is increased after consensus is achievedThe rate is mu; setting the release time of the transaction a as T ₁ Parasitic chain start time T ₂ The method comprises the steps of carrying out a first treatment on the surface of the If T ₂ After the consensus completion time t for transaction a, then transaction b will verify transaction a, i.e., each transaction verifying transaction b will indirectly verify transaction a, and parasitic link attacks are unlikely to succeed, so the parasitic link start time should be less than the consensus completion time t ₀ The method comprises the steps of carrying out a first treatment on the surface of the When the trade a consensus is reached, the cumulative weight of the parasitic chain is μ (t ₀ -T ₂ ) The cumulative weight of honest nodes is phi (t ₀ -T ₁ ) The method comprises the steps of carrying out a first treatment on the surface of the Thus, if the attack is successful, after consensus is reached, the parasitic chain cumulative weight should be greater than that of the honest node, i.e., φ (t ₀ -T ₁ )-μ(t ₀ -T ₂ )<0; setting T ₁ ＝T ₂ =0, i.e. the parasitic chain starts to be ready when transaction a writes to log, then phi (t ₀ )-μ(t ₀ )<0; the cumulative weights phi (t) at the completion of consensus are calculated respectively ₀ ) The preparation is started at the moment of writing a log with the parasitic chain transaction a, and the probability of attack success is increased;

cumulative weight at transaction consensus completion: when a new transaction arrives, at least one verification site i exists in the k selected tips, and the new transaction indirectly verifies the site i, wherein the accumulated weight of the site i is added with 1; at least one tips verifies that the probability of site i is 1- (1-psi (t-h)/R ₀ ) ^k

Wherein phi (t) is the cumulative weight;

the method can obtain:

wherein,

by the nature of the omega functionThe method can obtain:

thus, the cumulative weight φ (t) can be calculated as:

since the weight accumulation time is mainly in the data transmission stage and the consensus achievement stage, t in the above formula is the sum of the data transmission time and the accumulated weight time, t ₀ The cumulative weight is expressed as follows:

probability of success of parasitic chain attack: because the honest nodes accord with poisson distribution with the event arrival rate lambda, the time intervals of adjacent things of the honest nodes are index variables which are independently and uniformly distributed and obey 1/lambda index distribution; similarly, the time interval between adjacent things of the malicious node is 1/mu of independent index variable distributed in the same way; the method can obtain:

when mu>Lambda, when mu is less than or equal to lambda, only when the transaction a agrees, the accumulated weight of the transaction b is greater than that of the transaction a, so that the attack is possible to be successful, namely phi (t ₀ )-μ(t ₀ )>0; upon successful attack, the malicious node should track the difference φ (t) of the cumulative weights of the parasitic chain and the validation transaction a ₀ )-μ(t ₀ ) Therefore, the probability of success of a parasitic chain attack is:

4. the method of claim 1, wherein the establishing a blockchain consensus algorithm based on the field network and the IOTA in step S3 specifically includes: forming a cluster by an IOTA node and adjacent terminals; the terminal sends a transaction to the IOTAnode; when the IOTA node receives enough transactions, it will pack all the transactions and assign PoW tasks to the terminals in the cluster; all members of the cluster cooperate with each other to complete respective tasks, and the calculation result is returned to the IOTA node, and then the IOTAnode writes data into a chaotic node; thus, the amount of IOTAnode effort and the time required to write data to chaotic data can be greatly reduced;

after clustering, after waiting for the transaction to reach the maximum value of the IOTA packet, the data can be written into the log, and the data collection time is as follows:

wherein n is _t N is the average transaction number in one data packet _g Lambda is the number of nodes in a cluster _i Event arrival rate for node i; the time of writing data into the table is mainly the PoW calculation time, and after clustering, poW is completed cooperatively by members in the cluster, so that the time of writing data into the table is as follows:

Wherein d _Pi For counting cycles of PoW calculation, F _i Computing power for cluster iThe average time interval for the event to write to the local log is:

because the data only needs to be transmitted to the IOTAnode node in the network, the transmission time of the data in the network is less than or equal to the transmission time when the data is not clustered, and the maximum transmission time LT is taken here;

the consensus achievement time is:

wherein->The consensus time is expressed as follows: