CN103200616B - A kind of energy-efficient deployment method setting up Internet of Things network model - Google Patents

Abstract

The invention discloses the method for the foundation of a kind of network model based on technology of Internet of things and energy-efficient deployment, belong to Internet of Things communication technical field.The present invention proposes the three-layer network framework based on technology of Internet of things and network energy consumption Optimized model.First the energy-efficient deployment method that the present invention proposes gets rid of the alternative via node that all can not communicate with any sensing node according to sensing node and the distance of alternative via node, then in remaining via node, i is selected, if they meet " all standing sensing node and whether be connected graph with the topology that base station is formed ", the sum that then sending node of every section of link transmission power consumption and the reception of receiving node consumed energy is equivalent to the weight of this section of link, the minimum value of energy consumption and the life span of network is obtained by the Dijkstra method improved, using network node deployment way corresponding under this energy consumption as a kind of maximum energy-saving deployment way.

Description

A kind of energy-efficient deployment method setting up Internet of Things network model

Technical field

The present invention relates to Internet of Things communication technical field, especially set up and energy-efficient deployment policing issue for the network model based on technology of Internet of things.

Background technology

Along with the develop rapidly of the communication technology, people have no longer been satisfied with interpersonal and have needed people to participate in mutual communication mode, one is more intelligent, more easily communication mode expected by people.Internet of Things---the communication mode that a kind of object, machinery compartment do not need the participation of people can complete information interaction just arises at the historic moment.Internet of Things (InternetofThings) is a kind of intelligent network connected together by all objects, utilize as radio-frequency (RF) identification (Radio-FrequencyIDentification), radio communication, in real time location, Video processing and sensing technology and equipment, make any intelligent object carry out information interchange through network.By RFID(radio-frequency (RF) identification) based on system, in conjunction with existing network technology, sensor technology, database technology, middleware Technology etc., form the trend that a huge network be made up of a large amount of read write lines and mobile tag becomes Internet of Things development.

But in real life, the sensing node of Internet of Things is mostly deployed in the scene of no worker monitor, have ability fragility, the feature such as resource-constrained, often have adding or the inefficacy of existing node of new node, topology of networks change is very fast, network is once be formed, and people seldom intervenes it and runs.And the wireless communication bandwidth of sensing node is limited, usually only have the speed of hundreds of kbps, battery capacity is not generally very large, its special application determines in use, can not charge the battery or change battery, once the energy content of battery is finished, this node also just loses effect.In addition, country is advocating energy-saving and emission-reduction work always in recent years, and therefore in Internet of Things design process, the use of any technology and agreement all will premised on energy-conservation.

At present, Internet of Things is just in the elementary step of research, a not unified framework, more ununified standard can be followed, also little to the correlative study based on power saving in technology of Internet of things, but more existing researchers propose relevant models and methods for the power saving in wireless sense network.The people such as professor Wang Xinbing study by equivalent sensing radius the WSN(WirelessSensorNetwork that sensor node is Uniformly distributed and Poisson distribution) the heterogeneous and mobility of interior joint is on the impact of network in spreadability and energy ezpenditure, the mobility obtaining egress by experiment can reduce perception power consumption, the heterogeneous energy ezpenditure adding one-dimensional random mobility model on the contrary, does not affect independent same distribution model; The people such as KonstantinosOikonomou have studied link weight and change along with the change of the energy level between adjacent node and link load, meeting point can move according to the extendible meeting point migration strategy of one between neighbor node simultaneously, and this dynamic meeting point scheduling mode can reduce network energy consumption, extend network time.The people such as OnurSoysal have studied stochastic route mechanism energy-conservation based on MAC (MediumAccessControl) layer in WSN, propose a kind of can the agreement POWERNAP of adjustment node sleep-awake mechanism, calculate the overhead that schedule information that its sleep-awake dispatch list solves a large amount of complexity is brought to node by making receiver, thus save the total energy consumption of system.Although the research in the past about radio sensing network energy consumption problem is optimized network energy consumption, but mostly only consider the situation that simple power consumption constraint or network lifetime retrain or combine both consideration, but the constraint of the cost of system and link flow Constraints of Equilibrium are also factors very important in Internet of Things.Simultaneously, because the transmission radiuses of factor to route such as electromagnetic interference, air humidity, temperature in wireless sense network have a great impact, Dynamic routing mechanisms almost can not be applied out of doors, and all sensing nodes can perception data and mutual forwarding data in traditional AdHoc network, bearing overweight load with regard to being easy to cause from the sensing node close to base station like this, finally causing network paralysis.

Summary of the invention

Problem to be solved by this invention is: large for Internet of Things scale, the characteristic of network node performance fragility, a kind of network architecture of layering is proposed, the energy ezpenditure constraint that this framework lower node of comprehensive study causes when sending data, link flow Constraints of Equilibrium, link maximum transmission rate constraint, system budget constraint and network lifetime evaluation criterion, solve traditional simple research energy consumption problem or life span or the two and combine the problem that the overall performance of network caused can not reach optimum, and a kind of static routing optimized algorithm being applicable to Internet of Things is proposed, solve the problem that energy loss in technology of Internet of things is too much, avoid electromagnetic interference, air humidity, the factors such as temperature cause Dynamic routing mechanisms can not be applicable to extensive Internet of Things well.

The scheme that the present invention solves the problems of the technologies described above is, a kind of energy-efficient deployment method setting up Internet of Things network model, comprise step, the Internet of Things network model building a kind of layering comprises sensing layer, relay layer and convergence-level, the task different in a network according to each node layer, sensing layer node is responsible for perception and is sent data, perception data is not directly passed to convergence-level, but the base station of convergence-level is transmitted to by the via node of relay layer, perception data is not transmitted mutually between sensing node, by to divide the work layer by layer like this and co-operating mode reaches the object optimizing Internet of Things.The network architecture for this layering proposes a kind of Optimized model, this Optimized model comprehensive study energy ezpenditure constraint, link flow Constraints of Equilibrium, system budget constraint and network lifetime evaluation criterion.To propose in a kind of algorithm calculation optimization model four kinds of constraints value and between impact, relatively draw a kind of energy-efficient deployment mode, it is characterized in that, the quantity of first given sensing node, base station and alternative via node and position, calculate the distance of each sensing node and alternative via node, remove the alternative via node all do not communicated with any sensing node, secondly from remaining alternative via node, select and satisfy condition i the via node of " can all standing sensing node and be connected graph with the topology that base station is formed ", according to formula determine that the sensing node unit interval sends power consumption, according to formula determine that the via node unit interval receives power consumption, according to formula determine that the via node unit interval sends power consumption, according to formula determine that the basic unit time receives power consumption, and the reception of the transmission power consumption of every section of both link ends sending node in topological diagram and receiving node power consumption sum is equivalent to the weight of this section of link, then the dijkstra's algorithm of improvement is utilized to find out the shortest path of base station to each sensing node, the length of all shortest paths obtained is added the energy consumption as current network topology, and according to formula obtain the life span of this kind of network topology, if the maximum link flow value flowing exceed regulation of certain section of link in the dijkstra's algorithm improved, then select suboptimum shortest path to realize the flow equilibrium constraint of link further, finally find the energy consumption of i all possible network topology of via node, find out minimum energy dissipation value, using via node, sensor node and the quantity of base station selected by corresponding for this least energy consumption and the deployment of the position energy-efficient deployment mode as Internet of Things.

The present invention is directed in Practical Project the problem of disposing large-scale sensor net and running into, a kind of layered network model being applicable to Internet of Things is proposed, to be divided the work layer by layer by node and co-operating mode reaches the object optimizing Internet of Things load, and by carrying out link flow Constraints of Equilibrium to model, link maximum transmission rate constraint, solves the problem that network that traditional AdHoc nodes laod unbalance causes is paralysed too early further.Meanwhile, four major constraints that comprehensive study Internet of Things is optimized, solve traditional simple research energy consumption or life span or the two and combine the problem that the overall performance of network caused can not reach optimum.Finally, show that a kind of node deployment mode of the best makes network reach the effect of energy-efficient deployment by optimized algorithm, because algorithm uses static routing mechanism, the factors such as electromagnetic interference, air humidity, temperature that avoid cause dynamic routing cannot the situation of large-scale application out of doors.

Accompanying drawing explanation

Fig. 1 is the three-layer network topology diagram based on technology of Internet of things of the present invention

Fig. 2 is Optimized model figure of the present invention

Fig. 3 is algorithm flow chart of the present invention

Embodiment

First we make following simplified characterization without loss of generality for Internet of Things network model:

1) represent two points in Euclidean plane with x and y, represent the distance between 2 with d (x, y);

2) with S set={ SN ₁, SN ₂..., SN _lrepresent given l the SN(sensor node be distributed in two dimensional surface), its communication radius is designated as r>0;

3) with set χ={ RN ₁, RN ₂..., RN _mrepresent m the RN(via node being distributed in two dimensional surface), its communication radius is designated as R>=r;

4) with set B={ BS ₁, BS ₂..., BS _nrepresent n the BS(base station be distributed on two dimensional surface), anticipate between two base stations and all can intercom mutually.

For accompanying drawing, network model proposed by the invention is illustrated below.

Figure 1 shows that the three-layer network topology diagram based on technology of Internet of things of the present invention.

A kind of layered network architecture based on Internet of Things, Internet of Things interior joint is distributed in sensing layer, relay layer and convergence-level by the difference according to node tasks, sensing layer mainly comprises some awareness apparatus such as RFID (radio-frequency (RF) identification), transducer, GPS (global positioning system), video camera, laser scanner, relay layer mainly comprises the stronger via node of some functions, convergence-level mainly comprises the base station receiving perception information, and between each node of three-layer network, communication feature is described below:

1. for any SN _i∈ S, RN _j, if there is d (SN in ∈ χ _i, RN _j)≤r, then SN _ican to RN _jtransmission information;

2. for any RN _i∈ χ, N _j, if there is d (RN in ∈ χ ∪ B _i, N _j)≤R, then RN _ienergy and N _jcommunication;

3. for arbitrary SN _i∈ S, SN _j, even if there is d (SN in ∈ S _i, SN _j)≤r, SN _iand SN _jcan not intercom mutually.

Optimized model figure of the present invention shown in Fig. 2.Illustrate for several constraints in model below.

An Internet of Things network diagram is represented with G (V, E), wherein, the nonempty set that V is made up of network node, | V| represents the quantity of nodes, and E is the set of wireless link in network.Because sensing node in three-layer network model is only responsible for sending data to via node, and via node sends data to base station and adjacent via node, so this three-layer network figure is a directed connected graph.If can communicate between node i with node j, then title node j is the adjacent node of node i.N (i) represents the adjacent node set of node i.Represent internodal relation with network connection matrix A, the element definition in matrix is:

$a_{\mathrm{ij}}=\left\{\begin{array}{cc}1,& j\∈N\left(i\right)\\ 0,& j\∉N\left(i\right)\end{array}\right.---\left(1\right)$

A _ijfor element in network connection matrix, a _ijwhether decision node i and node j adjoins.

The present invention considers following four kinds of constraintss, proposes a kind of Optimized model:

1) energy ezpenditure constraint

Because the energy ezpenditure of data communication is much larger than data induction and the energy ezpenditure of data processing, therefore the energy ezpenditure of data communication only considered by this model, i.e. the energy ezpenditure that transmits and receive data of node, known by free space channel model:

Node sends data power consumption: E _tx=(E _elec+ ε d ²) L(2)

Node receives data power consumption: E _rx=E _elecl(3)

In formula, d is internodal distance, and L is data volume, E _elecelectric circuit electronics technical energy consumption during expression transmitting-receiving unit data, ε is and joint behavior parameter.

According to the feature of three-layer network model proposed by the invention, by above free space channel model, we can obtain the power consumption of each node in unit interval system respectively.

The energy consumption of unit interval sensing node i:

$e_i=\underset{j\∈\mathrm{RN}}{\mathrm{\Σ}}{a}_{\mathrm{ij}}{F}_{\mathrm{ij}}(E_{\mathrm{elec}}+{\mathrm{\ϵ}}_1{d_{\mathrm{ij}}}^2),\∀i\∈S---\left(4\right)$

The energy consumption of unit interval via node j:

$e_j=\underset{i\∈(\mathrm{SN}\∪\mathrm{RN})}{\mathrm{\Σ}}{a}_{\mathrm{ij}}{F}_{\mathrm{ij}}{E}_{\mathrm{elec}}+\underset{i\∈(\mathrm{BS}\∪\mathrm{RN})}{\mathrm{\Σ}}{a}_{\mathrm{ji}}{F}_{\mathrm{ji}}(E_{\mathrm{elec}}+{\mathrm{\ϵ}}_2{d_{\mathrm{ji}}}^2),\∀j\∈\mathrm{\χ}---\left(5\right)$

The energy consumption of unit interval base station k:

$e_k=\underset{j\∈\mathrm{RN}}{\mathrm{\Σ}}{a}_{\mathrm{jk}}{F}_{\mathrm{jk}}{E}_{\mathrm{elec}},\∀k\∈\mathrm{BS}---\left(6\right)$

ε in formula ₁, ε ₂represent sensing node and via node hardware performance parameter respectively, E _elecelectric circuit electronics technical energy consumption during expression transmitting-receiving unit data, F _ijfor node i sends data to the data transfer rates of j, d _ijrepresent the distance between node i and node j, RN, SN, BS represent the set of via node, sensor node, base station respectively.

2) link flow Constraints of Equilibrium

Due to the limited bandwidth resources of node-node transmission, the data transmission bauds on link is also limited, i.e. link maximum transmission rate constraint.In three-layer network model of the present invention, can receive mutually and send data between via node, so the link formed between via node will meet following constraint:

$a_{\mathrm{ij}}{F}_{\mathrm{ij}}+a_{\mathrm{ji}}{F}_{\mathrm{ji}}\≤H_{\max },\∀i,j\∈\mathrm{\χ}---\left(7\right)$

H in formula _maxfor link maximum transmission speed, F _ijfor node i sends data to the data transfer rates of j, χ refers to the set of via node.

Equally, in three-layer network model, only have sensing node to send data to via node between sensing node and via node and the link formed, only have via node to send data to base station between via node and base station and the link formed, retrain below these link demand fulfillment:

$a_{\mathrm{ij}}{F}_{\mathrm{ij}}\≤H_{\max },\∀i\∈S,j\∈\mathrm{\χ};i\∈\mathrm{\χ},j\∈B---\left(8\right)$

H in formula _maxfor link maximum transmission speed, S, χ, B refer to the set of transducer, via node, base station respectively.

3) system budget constraint

In Internet of Things, because the cost of base station and via node is relatively high, so we can dispose a small amount of via node and base station reaches the object reducing system cost budget as far as possible, its sensor node, via node and base station meet following relation.

0<C _S|S|+C _R|χ|+C _B|B|<W ₀（9）

C in formula _s, C _r, C _bbe respectively the unit price of sensing node, via node and base station, | S|, | χ | with | B| represent respectively to dispose sensor node, via node, base station quantity, W ₀for system maximum budget.

4) network lifetime

Network lifetime is one of most important index in Internet of Things.In Adhoc net network lifetime is defined as network to bring into operation this period of time that the 1st node energy in network exhaust, similar, network lifetime is defined as this period of time of the 1st via node depleted of energy in the network operation to network by the present invention according to the feature of three-layer network model, therefore the energy consumption of network lifetime T and via node j and the primary power E of via node ₁meet relation:

$T=\min \left\{\frac{E_1}{e_j}\right\},\&ForAll;j\&Element;\mathrm{\&chi;}---\left(10\right)$

Namely $e_j\&le;\frac{E_1}{T},\&ForAll;j\&Element;\mathrm{\&chi;}---\left(11\right)$

Wherein e _jtried to achieve by formula (5).

5) minimum energy dissipation of network is as target function

The present invention mainly studies the power consumption issues in Internet of Things, and target function is the minimum energy dissipation of network: $\min (\underset{i\&Element;\mathrm{SN}}{\mathrm{\&Sigma;}}{e}_i+\underset{j\&Element;\mathrm{RN}}{\mathrm{\&Sigma;}}{e}_j+\underset{k\&Element;\mathrm{BS}}{\mathrm{\&Sigma;}}{e}_k)$

The various constraintss of comprehensive study Internet of Things of the present invention, propose Network Optimization Model.A kind of energy-efficient deployment mode is obtained by the quantity of the network node corresponding to the minimal network energy consumption obtained and position, and in the process of Algorithm for Solving least energy consumption, consider link flow Constraints of Equilibrium, the further offered load that improves is to extend network lifetime, consider the budgetary restraints of system simultaneously, reduce system budget by the deployment quantity of constraint network node, finally carry out the performance of overall merit network in conjunction with network lifetime.

It is algorithm flow chart of the present invention shown in Fig. 3.The energy-efficient deployment method that the present invention proposes comprises: the distance calculating each sensing node and alternative via node, remove the alternative via node all do not communicated with any sensing node, then in remaining via node, i is selected, judge that they whether can all standing sensing node and whether be connected graph with the topology that base station is formed, if meet this two conditions simultaneously, the sum that every section of link sending node transmission power consumption and the reception of receiving node consumed energy is equivalent to the weight of this section of link, calculate the minimum value of energy consumption, using network node deployment way corresponding under this energy consumption as a kind of energy-efficient deployment mode.Illustrate for the several key steps in algorithm below.

Step 1: according to the position coordinates of sensor node in Internet of Things and possible via node, calculate each possible via node RN _jwith each sensor node SN _idistance d _ijif, d _ij<r(r is the communication radius of sensor node), just by this via node RN _jadd alternative set of relay nodes to, simultaneously by this sensing node SN _ijoin this via node RN _jabutment points set N (RN _j).

Step 2: selecting all possible i(i from alternative set of relay nodes is the integer being less than alternative via node quantity) individual via node, and ask the abutment points union of sets collection ∪ N of this i via node _i(RN _j), wherein N _i(RN _j) represent the abutment points set of i-th alternative via node selected, judge ∪ N _i(RN _j) whether equal sensing node set, if equal, represent that selected via node can all standing sensing node, judge whether the network topology that this i via node and base station are formed is communicated with according to depth-first search again, if this i via node selected is an all standing of sensing node, and the network diagram of the via node selected and base station formation is connected graph, each sensing node then can be made to send data to base station, otherwise then can not.

Step 3: obtain via node number and be i and after all possible network topology of satisfy condition " can all standing sensing node and be connected graph with the topology that base station is formed ", for each network topology, the power consumption values of every category node can be obtained according to formula (3) (4) (5), because three-layer network topology is a directed graph, the weight of every section of link is equivalent to sending node and sends power consumption and the reception of receiving node and to consume energy sum by the present invention, find out the shortest path (can utilize the dijkstra's algorithm of improvement) of base station to each sensing node, the length of all shortest paths obtained is added the energy consumption as current network topology, be the power consumption of wherein a kind of network topology when via node number is i, and the life span of this kind of network topology is obtained according to formula (10), if the maximum link flow value flowing exceed regulation of certain section of link in the dijkstra's algorithm improved, then select suboptimum shortest path to realize the flow equilibrium constraint of link further.

Step 4: the power consumption of all possible network topology when via node number obtained in comparison step 3 is i, draw energy consumption when minimum value is i as via node number, using the energy-efficient deployment mode that sensor node corresponding under this energy consumption, base station and the quantity of via node selected and position are carried as the present invention.

Claims (3)

1. set up the energy-efficient deployment method of Internet of Things network model for one kind, it is characterized in that, build the Internet of Things network model comprising sensing layer, relay layer and convergence-level, sensing layer node is responsible for perception and is sent data, perception data is transmitted to the base station of convergence-level by the via node of relay layer, between sensing node, does not transmit perception data mutually, calculate the distance of each sensing node and alternative via node, remove the alternative via node all do not communicated with sensing node, select from remaining alternative via node and satisfy condition: i the via node of " can all standing sensing node and be connected graph with the topology that base station is formed ", to arbitrary via node wherein, according to sensing node unit interval energy consumption, via node unit interval energy consumption, basic unit time energy consumption, and the reception energy consumption sum of the transmission energy consumption of every section of both link ends sending node in topology and receiving node is equivalent to the weight of this section of link, find out the shortest path of base station to each sensing node, the length of all shortest paths is added the energy consumption as current network topology, according to formula: determine sensing node unit interval energy consumption, according to formula: determine via node unit interval energy consumption, according to formula: determine basic unit time energy consumption, call formula thus the least energy consumption of computing network, is deployed as according to the quantity of the via node selected by least energy consumption correspondence, sensor node and base station and position the Internet of Things meeting energy-efficient deployment, wherein, and ε ₁, ε ₂represent sensing node and via node performance parameter respectively, E _elecelectric circuit electronics technical energy consumption during expression transmitting-receiving unit data, F _ijfor node i sends data to the data transfer rates of j, d _ijrepresent the distance between node i and node j, RN, SN, BS represent the set of via node, sensor node, base station respectively, a _ijfor element in network connection matrix.

2. energy-efficient deployment method according to claim 1, is characterized in that, further, the link formed between via node meets: $\begin{array}{cc}{a}_{\mathrm{ij}}{F}_{\mathrm{ij}}+a_{\mathrm{ji}}{F}_{\mathrm{ji}}\&le;H_{\max }& \&ForAll;i,j\&Element;\mathrm{\&chi;}\end{array},$ Sensing node sends data to link, via node that via node formed and sends data to the link that base station formed and meet: i ∈ χ, j ∈ B, wherein, H _maxfor link maximum transmission speed, F _ijfor node i sends data to the data transfer rates of j, a _ijfor element in network connection matrix, meet: $a_{\mathrm{ij}}=\left\{\begin{array}{cc}1,& j\&Element;N\left(i\right)\\ 0,& j\&NotElement;N\left(i\right)\end{array}\right.,$ S represents the sensor node set be distributed in two dimensional surface, and χ represents the set of relay nodes be distributed in two dimensional surface, and B represents the collection of base stations be distributed in two dimensional surface.

3. energy-efficient deployment method according to claim 1, is characterized in that, further, the unit interval energy consumption of via node j meets: $\begin{array}{cc}{e}_j\&le;\frac{E_1}{T}& \&ForAll;j\&Element;\mathrm{\&chi;}\end{array},$ Wherein, T is network lifetime, E ₁for the primary power of via node.