1. bookAHEAD OF PRINT
Détails du magazine
License
Format
Magazine
eISSN
2444-8656
Première parution
01 Jan 2016
Périodicité
2 fois par an
Langues
Anglais
Accès libre

Information transmission simulation of Internet of things communication nodes under collision free probability equation

Publié en ligne: 15 Jul 2022
Volume & Edition: AHEAD OF PRINT
Pages: -
Reçu: 30 Apr 2022
Accepté: 30 Jun 2022
Détails du magazine
License
Format
Magazine
eISSN
2444-8656
Première parution
01 Jan 2016
Périodicité
2 fois par an
Langues
Anglais
Introduction

In recent years, VLSI (very large scale integration) and MEMS (micro-electro-mechanical systems) technologies have developed rapidly. This makes it possible to deploy international wireless sensor networks (WSNs) with a large number of small nodes [1]. WSNs nodes composed of sensors with sensors, information processing capabilities, and short-term wireless communication capabilities have broad application prospects in military defense, environmental monitoring, target tracking, biomedicine, emergency disaster relief, and commercial use [2]. Compared with other typical wireless networks, WSNs are also quite different. Although WSNs are indeed the classic representative of American Mobile Ad Hoc Networks (MANETs), they are quite different from MANETs themselves. WSNs nodes usually do not move, or move very rarely [3]. Because the data packet size of WSNs is smaller. The computing, storage, and communication capabilities of WSNs nodes are also limited. Therefore, WSNs are generally networked separately. It is mainly used for monitoring functions. It is also a data-centric networking system. Although the number of WSNs nodes can be as many as thousands. But it is much larger than dozens of nodes of MANETs [4]. WSNs are usually quite cumbersome and even impossible to replace batteries on installed nodes [5]. This also leads to optimal network survivability, which is the main design goal of WSNs. Due to the characteristics of WSNs, the MAC (mediu maccess) technology of many fixed networks and MANETs cannot reasonably use WSNs.

In this paper, according to the basic characteristics of WSNs network, several classic MAC protocols are expounded. On this basis, we provide a schedule-based sensor MAC (SSMAC), a MAC protocol for WSNs based on reservation scheduling. This protocol can provide protection of QoS service (quality of service) of data while saving energy. We can use the simulation analysis to obtain the characteristics of the SSMAC protocol.

Related work

The energy consumption in wireless sensor network system is generally manifested in many important technical aspects such as collision, crosstalk, control packet overhead, idle listening and premature transmission [6]. However, the MAC protocol in the current sensor network system has achieved the purpose of reducing energy by restricting the energy consumption in one or several ways [7]. On the other hand, in the structure of the sensor computer network, there is a kind of sink node. The node usually has a higher ability to transmit and process data. It produces more energy. And the ability to send data back to a remote processing center [8]. While other nodes adopt more hop relay method to send the received data to sink node. Therefore, the communication modes of nodes in WSNs can also include broadcast, centralized transmission, local communication and multicast communication [9]. The MAC protocol should also support this communication mode with lower energy consumption. It is also convenient for nodes to cooperate with each other to complete the tasks of monitoring and data communication [10].

The goal of the MAC protocol is to minimize network lifetime and protocol expansion. This change mainly involves the scale and topology of the network, the density of nodes, the addition of new nodes, and the demise of old nodes [11]. In addition, network characteristics such as throughput, delay and bandwidth utilization are also indicators that need to be considered. At present, researchers have designed many MAC protocols to serve WSNs. We can classify MAC layer agreements into contention establishment and reservation scheduling according to the characteristics of the agreement [12]. So, people provide a S-MAC (sensor-MAC) which can be carried out by 802.11DCF mechanism. The S-MAC protocol divides the clock into frames. The frame length is determined by the application [13]. The frame includes the working process and the sleep phase. It reduces idle listening time by periodically listening and sleeping with a stable time. It reduces conflicts of interest by adopting RTS/CTS/DATA/ACK methods. It reduces packet transmission costs by employing message passing. However, the broadcast data message cannot use the RTS/CTS method. This increases the chance of conflict. In addition, crosstalk also consumes some energy [14]. A fixed-length working time can impair the reliability of the protocol even under the premise of variable load. The main purpose of the T-MAC (timeout-MAC) we provide is to improve the characteristics of the S-MAC protocol under variable load, and to dynamically adjust the frame length of the working process to meet the needs of the variable load. However, it also raises a kind of “early bed” question. It is called early sleep, which means that when several sensor nodes transmit data information to a single or a few sink nodes, the nodes will not be triggered by events and fall into sleep. At this time, the node will not be able to detect the next data transmission, and there will be a communication delay. Therefore, the WiseMAC protocol must use non-persistent CSMA to transmit data packets. Idle nodes can periodically monitor the channel for short periods of time while determining the current channel condition [15]. In addition, the WiseMAC protocol can also use the data acknowledgment packet to carry the next channel monitoring time of the node. It adds a wake-up preamble combined with the clock skew rate before sending out a data frame. In this way, the connecting node can join the normal working condition and receive the data before the various data information parts in the frame are sent out. However, the short channel listening time stored in all adjacent nodes will waste precious space, thus increasing the complexity of the protocol implementation. Especially in the network system with high node density coefficient, this kind of problem is particularly obvious. Non-persistent, CSMA does not overcome the contradictions caused by hidden terminals. The Data-gathering MAC (D-MAC) agreement is an improved agreement of a time slot Aloha. Time slots are assigned to a tree from sensor nodes to sink nodes to form a data information aggregation tree. The data information in the tree is one-way. The data information message is sent from the child node to the parent node. Working/sleep state transition technology is used between nodes [16]. This makes the forwarding time of the child node overlap with the acceptance time of the parent node. In the best case, the forwarding of data information will be completed continuously without any delay. Because D-MAC cannot adapt to the situation of two-way data communication in the Internet, it cannot realize all the communication modes in WSNs. The TRAMA (traffic-adaptive MA Cprotocol) agreement refers to the technological progress in the field of energy saving based on the traditional TDMA agreement. It divides time into alternating random visit times and scheduled visit times. It uses the messages of all neighbor nodes in the local two-hop, and uses the allocation election mechanism to determine the conflict-free message senders in all time slots. Also, its use avoids allocating all slot messages to nodes with no traffic. It keeps the non-transmitting and receiving nodes in a dormant state to achieve the purpose of saving power. At the same time, the TRAMA protocol increases the network throughput, and also solves the defect of poor scalability of the MAC protocol using TDMA. However, the random use period of the TRAMA agreement is correspondingly longer. It occupies at least 12.5 percent of the entire time frame. However, due to the complicated time distribution election mechanism, the large data packet waiting in the queue has a large delay. Therefore, people generally think that the TRAMA protocol is more suitable for use scenarios that are not sensitive to delay.

In the Adhoc network system, there are many classic MAC layer protocols used in the synchronization network system. Such as FPRP (five-phase reservation protocol) and E (evolutionary)-TDMA and so on. FPRP is a set of distributed reservation adjustment mechanisms that can use five-way handshakes for high-probability and non-contradiction within the two-hop area. It is also a broadcast regulation protocol. But once the unicast is adjusted, its effect will not be high. The E-TDMA is a modified version of FPRP used. It, like FPRP, requires a five-way handshake to connect between nodes. Therefore, although this agreement can ensure the transmission of real-time services without conflict, it can also prevent delay jitter. But its control cost is very high and the realization process is complicated. So the efficiency of this method is not high. The use of the scheduled MAC protocol can ensure that the nodes transmit data packets without contradiction under the condition that the agreed tasks are completed. This contract property depends on the cost of the reservation calculation. At the same time, it can also obtain QoS protection. This paper will design a scheduling MAC protocol based on energy constraints and differentiated services in WSNs. This aspect overcomes the traditional hidden port and exposed port problems. In addition, we also fully consider the energy-saving features and have less multimedia connection delay, because these can provide some QoS protection for the data.

Agreement

The SSMAC protocol can be applied to the wireless sensor network in which the whole network synchronously divides time slots. There may also be two implementation methods in the system at the same time. a) Use GPS receiver to obtain certain precise frequency criterion in the air. It can work with controlled moments. Nodes are also able to obtain time synchronization based on a moment referenced to Coordinated Universal Time (UTC). b) Use the method of automatic time calibration to perform time synchronization with a certain accuracy. Since the data transmission rate of the sensor network is relatively small, the duration of each time slot is much larger than the typical time offset. At the speed of 115.2kbps, we can send a data packet of 512 Bytes using a time slot of up to 46 ms. At the same time, we can also perform node synchronization in WSNs by using a simple timestamp mechanism. The SSMAC protocol adopts a high-probability, non-contradictory distributed reservation scheduling management mechanism with multiple handshakes and two-hop range. It implements the QoS of the service according to the requirement of capability restriction and fully considering the support of the real-time service.

Channel division

At the beginning of the booming development of wireless network, it was deeply influenced by the hardware and software environment at that time. It mainly utilizes a single channel. Therefore, some classical and practical MAC protocols have been further developed. In the civil field, the most commonly used is the IEEE 802.1 series. While the traditional TDMA classification method is simpler to implement in hardware when it solves the synchronization problem of the whole network, it becomes the most effective classification method.

As shown in Figure 1, Each scheduling connection component of a time frame has K information frames. And each information frame contains N message time slots. The soft setting part in the random connection is generally used for the node that successfully reserved the message time slot last time. According to the specific calculation method, there is no need to compete for the connection. And the message time slot with the same sequence number is reserved again. In the connection phase of each competition, it will determine how nodes agree to schedule information frames. The appointment process includes N appointment time slots. It will in turn appoint service time slots with the same sequence numbers. And each successful appointment will make it occupy K relevant information time slots with the same sequence number in the subsequent scheduling access channel. Each appointment time slot will include M1+M2 appointment cycles (Figure 2). Its main purpose is to reduce the impact on the terminal environment and avoid exposing terminal problems. In this way, the Internet can carry larger parallel transmission of information and increase the success rate of reservation.

Figure 1

Time frame diagram

Figure 2

Schematic diagram of reservation time slot

How the protocol works

The TDMA protocol based on the principle of dynamic allocation focuses on two aspects of channel distribution and connection management. The former is generally responsible for relatively distributing appropriate channels for communication nodes. This enables as many nodes as possible to communicate simultaneously without conflict with each other. The latter is responsible for determining when nodes connect to channels, conflict resolution methods, and processing methods. Therefore, the SSMAC agreement basically completed the above two main functions. A channel allocation method in a TDMA network system is to assign transmission time slots to nodes in the network system. The collision-free transmission of packets between adjacent nodes is performed by means of repeated handshakes or agreements. This enables the highest possible wireless channel efficiency and space reuse. At the same time, the nodes that have nothing to do with this network system enter the dormant state to save power.

QoS Guarantee Mechanism

QoS is a kind of controllable pre-definition that users need to realize when transmitting end-to-end data information in computer network, according to the service attributes of end-to-end characteristics. It is usually divided into delay, delay jitter, bandwidth usage and packet loss rate. The prioritization of services and the reservation of resources are mainly used in two ways to ensure QoS in wireless networks. As long as the characteristics of the user service can meet the standard, it can be used as a QoS guarantee. The QoS guarantee requires multiple coordinations at each level to achieve business requirements. The business priority is mainly to classify the guaranteed services into the following categories. Usually we do it through different access priority levels and transmission probability. This ensures that high-priority services can occupy the channel with a high probability. The method of resource reservation is mainly to save some data of the channel. Only in this way can the channel use right of the constant bit rate and high priority service be ensured. Because then the real-time service can always provide the transmission channel.

The SSMAC protocol realizes QoS protection by means of differentiated services. The data packets generated by the upper layer are divided into two categories. Its contents are placed in two distinct data buffers in turn. These two kinds of data buffers are the real-time service area and the general service area in turn. When providing reservation service for data packets in the protocol, we will take completely different countermeasures for the number of packets in each buffer in our own queue. Assuming that the number of real-time service packets still exists in the buffer when the next time frame arrives, the software reservation mechanism is enabled to preempt time slots with the same sequence number. At the same time, the reservation process can be started only when the ordinary service meets the number of K data packets. At the same time, the number of packets in the cache must meet the corresponding number before the software reservation mechanism can be enabled.

Start the soft reservation mechanism

The header of each time frame has a soft reserved time slot with a length of N bits. It is sequentially relative to the next N normal message time slots. If the node sees the real-time ordinary message to be sent in the real-time service buffer after the new time frame is started. Or there are 3×K ordinary message packets to be sent. The node sets the bit of the soft reserved part of the corresponding sequence number to one. At this point, the node has been successfully soft-reserved. But if the bit seen by another node is set to one, the set message time slot is no longer reserved. It only receives the broadcast message from the successful soft reservation link from a predetermined cycle of the corresponding reservation time slot. In this way, it is determined whether it is its recipient. The purpose is to maintain the fairness of the access channel transmission between nodes. So we can allow the node to use the soft reservation for the third time to reserve the information time slot. After reaching three, the node can use the reservation service to complete the occupation of the slot space.

Reservation mechanism of four-way handshake

The SSMAC protocol adopts the basic idea of FPRP five-way handshake. It uses multiple handshakes to make it possible to reserve unicast, multicast and broadcast, and supports QoS. Its reservation process is usually only included in sites within the second-hop area. This is also a local process. It can also support channel multiplexing in space. The states that a node may have in the agreement are the transmit state (TX), the receive state (RX), and the sleep state (SL). And when the computer network system is initialized, the nodes are in the TX and RX states. When networking, the agreement generally selects the reservation process after the network system reaches a steady state. Figure 3 shows the packet format of the competition packet.

Figure 3

Contention message format

Reservation Request Phase (RR)

The node will be at i of the current reservation slot. When there are real-time service data packets or K general service data packets in the node cache, the node will request to reserve resources for use. At this time, it includes the following three situations: a) The node has used the soft reservation method and obtained the ownership of the current time slot. The node will broadcast the RR packet of the agreed resource usage requirement to one-hop neighbor nodes once. In the RR packet, it should contain the ID of the sending node, the ID of the destination node in the K data packets, and the flag bit. This means that the node acquires the ownership of this time slot through soft reservation. b) The node listens to the soft reservation part. When we see that this slot has been soft-reserved by other nodes. At this time, it respects the reservation of other nodes and joins the RX state, and listens for RR packets sent by other nodes. C) No node soft setting. Neither is the node in the slot that successfully reserved the slot from the previous appointment period. At this time, the node transmits an RR packet of a reservation request to the one-hop neighbor node with the connection probability PRR. The connection probability PRR is determined by the multi-hop pseudo-Baysian method. In the RR packet, it should contain the ID of the forwarding node or the ID of the destination node of the K data packets. The nodes that cannot send RR packets should switch to the RX state and start listening at this moment. The node that sends out the RR packet is called the Reservation Node (RN) in the technology.

The node is in part ii of the reserved slot. Therefore nodes also have their own time slot table. It records the utilization and reservation status of the current time slot between nodes and neighbors. At this time, once the data packet of the neighbor node needs to be forwarded, it will detect the status of each neighbor. Each neighbor destination node of the forwarded neighbor node is not in the RX state, and the RR packet is sent only when the destination node that wants to send the message data is not in the RX state.

Transmission Report Phase (TR)

Assuming that the node accepts two or more agreed request packets in phase 1, the node will detect the contradiction. At this time, in the appointment period, several appointment nodes need to compete for appointment service. When the node is in the transmission report phase of the I part of the appointed time slot, once it detects the contradiction of the RR group, the node will broadcast a contradiction and report the CR group in the second phase. Otherwise it is silent. When the node is in the transmission report stage of the ii part of the appointed time slot, once the contradiction of the RR group is detected, the node will also broadcast a contradiction and report the CR group. When a node accepts an RR packet only once, the node will analyze its content. However, only when it is one of the destination address fields in the RR packet and it is in an idle state, it can accept the opponent's contract application. Otherwise, CR packets will be broadcast simultaneously. After listening to the CR at this stage, the RN determines whether its RR is in contact with other recipients. Once the CR is not received, the RN confirms that the RR sent by it has been accurately accepted by all neighboring nodes. This RN node also becomes a transfer node (TN). RR/CR interaction solves the hidden port problem. The interaction of the reservation process between the two parts exploits the power of exposed nodes to be able to teleport in parallel with neighboring nodes. This may also increase spatial multiplexing.

Reservation Confirmation Phase (RC)

In this stage, the convention is established. In this state, the TN broadcasts a set of contract confirmation RC packets. And inform a hop adjacent node that the corresponding time slot has been agreed. And every time, the one-hop neighbor node that actually receives these RC messages knows that the time slot has been agreed. Non-transmitting nodes will modify their own time slot list. It will also add the source node ID of the RC packet to the corresponding slot in the slot list. It will also receive the message from the TN in the corresponding time slot of the service channel. At the same time, it will no longer compete for the slot.

Reservation Confirmation Stage (RA)

The node that sent the Reservation Confirmation RC packet received the Reservation Confirmation RA packet. And tell TN and TN's two-hop neighbors to succeed. Both the two hop neighbor nodes find that the node two hops away has successfully reserved resources. Assuming that the TN is not a connecting node, it also does not receive an acknowledgment packet for the reservation. It can be seen that TN is a closed node. Therefore, there is no need for TN to transmit messages.

When the node performs a reservation cycle after the four-way handshake phase, it will enter the next reservation cycle. Once a node has been successfully reserved, it will enter the SL state to save energy. Until it enters the reserved time slot ii, or enters the next reserved time slot. The possible states of the nodes after performing the above-mentioned complete reservation process include: all the nodes after the reservation is completed will be in the TX state in the corresponding information time slot. The destination node in its one-hop neighbor node will be in the RX state in the corresponding information time slot. The other nodes will be in the SL state in the corresponding information time slot. Only nodes that reach the TX state can complete data transmission in the corresponding information time slot.

Energy saving performance of the protocol

The SSMAC protocol uses a distributed scheduling connection method. The contradiction of its data message and Chongqing media appear in the random connection stage. When entering the stage of scheduling connection, then the node will switch itself between the states of TX, RX and SL according to the restrictions set by the time slot table. In this way, data message exchange is realized. In the random connection phase, the data packet conflict and Chongqing media are generally grouped through RR. This is generally limited by the multi-hop Pseudo-Baysian algorithm. It uses both allocation and discrete conflict methods. The application of this calculation in FPRP has been shown to be feasible. The packet widths of the four handshakes in the reserved time slot are all different. Therefore, the width of the TR, RC and RA messages can be shortened to only pass the two message fields of the message class and the source address. According to the requirements and changes of the superior service, we can also set the basic parameters of the SSMAC protocol. Modify the ratio of the instant frame occupied by the random access part to meet various types of network services.

Algorithms
The total energy consumption model of the system

Assume that n cognitive radio nodes are distributed in a circular area with radius R, and their distribution satisfies the Poisson distribution expected to be λ (λ is the number of nodes per unit area). Nodes use secondary communication power to realize multi-hop communication (intra-cluster communication power and inter-cluster communication power).

The initial energy of the nodes is the same as E0. When describing the total energy consumption model of the system, we need to make the following assumptions: (1) The communication radius of common nodes is r. The communication radius of the cluster head node is 2r. (2) There is no collision or transmission error when the cognitive nodes communicate. Cognitive nodes do not need to retransmit packets for various reasons. (3) There is a communicating primary user within the communication range of the cognitive node.

According to the assumptions, we can get the expected number E[N] = λS of cognitive nodes in a circular area with radius R. where S = πR2. Assume that the probability of common nodes becoming cluster heads among cognitive nodes is p, so the density of cluster heads is λ1 = . The density of ordinary nodes is λ0 = (1 − p)λ. Suppose E[Nc] and E[Lc] denote the expectation of the number of common nodes in each cluster and the expectation of the sum of the distances between common nodes and the cluster head node, respectively, under the condition that the clustering is completed. Then there are: E[ Nc|N=n ]=E[ Nc ]=λ0λ1 E\left[ {{N_c}|N = n} \right] = E\left[ {{N_c}} \right] = {{{\lambda _0}} \over {{\lambda _1}}} E[ Lc|N=n ]=E[ Lc ]=λ02λ13/2 E\left[ {{L_c}|N = n} \right] = E\left[ {{L_c}} \right] = {{{\lambda _0}} \over {2\lambda _1^{3/2}}}

N is a random variable representing the number of cognitive nodes in the circular area. Here we assume that the energy consumption of the cognitive node to receive and send unit data is Er and Es, respectively. We let Er = βEs, where β ∈ [0, 1] is called the data reception energy consumption coefficient. We found that the value of this parameter is 0.73 through the actual measurement of the IEEE802·11 network card with 2M bandwidth.

If the communication between two cognitive nodes with distance d requires at least [d / mr] or [[d / 2mr] nodes as relays. where m ∈ [0,1] is called the node density correction factor. It is defined as the following formula. m=1eλ/100 m = 1 - {e^{ - \lambda /100}}

Since n nodes will form np clusters, let Em denote the energy consumption of all ordinary nodes in transmitting unit data. E[ Em|N=n ]=npE[ Lc|N=n ]mr(Pt+Pr) E\left[ {{E_m}|N = n} \right] = np{{E\left[ {{L_c}|N = n} \right]} \over {mr}}\left( {{P_t} + {P_r}} \right)

Eh represents the maximum energy consumption of all cluster head nodes when transmitting unit data, there are: E[ Eh|N=n ]=np(E[ Nc ]+1)2R2mr(Pt+Pr)+npPr E\left[ {{E_h}|N = n} \right] = np\left( {E\left[ {{N_c}} \right] + 1} \right){{2R} \over {2mr}}\left( {P_t^\prime + {P_r}} \right) + np{P_r} Where Pt P_t^\prime represents the transmit power of the cluster head node. Since its emission radius is 2r, the relationship between emission distance and energy consumption of the free-space channel model is Pt4Pt P_t^\prime \approx 4{P_t} . npRr Indicates the energy required by the cluster head node to detect the occupancy of the primary user's communication channel from time to time.

Eh<Eh E_h^\prime < {E_h} . If the main user stops communication or the communication distance between two nodes is less than 2R and other situations. Therefore, the formula is expressed as the upper limit of energy consumption when the cluster head node transmits unit data. Let Et denote the total energy required to transmit a unit of data in a cognitive radio network, then: E[ Et|N=n ]=E[ Eh|N=n ]+E[ Em|N=n ]=E[ N ][ (1+β)(1p)2mrpλ+(4+β)Rmr+βp ]Pt=λπR2Pt[ (1+β)(1p)2mrpλ+(4+β)Rmr+βp ] \matrix{ {E\left[ {{E_t}|N = n} \right] = E\left[ {{E_h}|N = n} \right] + E\left[ {{E_m}|N = n} \right]} \hfill \cr { = E\left[ N \right]\left[ {{{\left( {1 + \beta } \right)\left( {1 - p} \right)} \over {2mr\sqrt {p\lambda } }} + {{\left( {4 + \beta } \right)R} \over {mr}} + \beta p} \right]{P_t}} \hfill \cr { = \lambda \pi {R^2}{P_t}\left[ {{{\left( {1 + \beta } \right)\left( {1 - p} \right)} \over {2mr\sqrt {p\lambda } }} + {{\left( {4 + \beta } \right)R} \over {mr}} + \beta p} \right]} \hfill \cr }

Taking the derivation of p at both ends of Equation (6) and setting the derivation formula to zero, the following formula can be obtained: cp3/2p1=0 c{p^{3/2}} - p - 1 = 0 Where c=4mrλβ/(1+β) c = 4mr\sqrt \lambda \beta /\left( {1 + \beta } \right) . Solving higher-order equations with respect to p, we get three roots. We discard the two imaginary roots, and the real roots are represented as follows: p=[ 8+123c27c2+4+108c236c+13c+33c8+123c27c2+4+108c23 ]2 p = {\left[ {{{\root 3 \of {8 + 12\sqrt {3c} \sqrt {27{c^2} + 4} + 108{c^2}} } \over {6c}} + {1 \over {3c}} + {3 \over {3c\root 3 \of {8 + 12\sqrt {3c} \sqrt {27{c^2} + 4} + 108{c^2}} }}} \right]^2}

The optimal cluster head probability p(β, m) can minimize the energy consumption in the process of transmitting unit data in cognitive radio network. Let R = 10m, r = 1m, Pt = 1, m = 1, β = 1 be the unit energy of the transmitted energy and the received energy. Use MATLAB simulation software to draw the relationship between network energy consumption and p and λ in unit time as shown in Figure 4.

Figure 4

The relationship between network energy consumption and p and λ per unit time

The minimum energy consumption per unit time in the network is not the case with the least number of cluster head nodes in the network. This is because nodes need to consume a certain amount of energy to send and receive data. And each cluster head manages (1 − p) / p member nodes on average. In unit time, the data packets sent by (1 − p) / p nodes are also received. Therefore, too many or too few members will affect the total energy consumption of the network per unit time.

Calculation of Double Threshold Values of DTACRA Algorithm

We assume that the communication distance between common nodes in the cluster and the cluster head is not greater than the communication radius r of common nodes. The transmission and reception energy consumption of the communication node can be expressed as the following formula: { Er=EelectkEt=Eelectk+Eamplak \left\{ {\matrix{ {{E_r} = {E_{elect}}k} \hfill \cr {{E_t} = {E_{elect}}k + {E_{amp}}{l^a}k} \hfill \cr } } \right. Where Eelect is the circuit energy consumption. Eamp is the RF gain. k is the number of bits sent or received. l is the communication distance. α is the channel attenuation factor. Free space channel model α = 2. Therefore formulas (4) and (5) can be further formulated as follows: E[ Em|N=n ]=npkE[ Nc ](2Eelect+Eampr2) E\left[ {{E_m}|N = n} \right] = npkE\left[ {{N_c}} \right]\left( {2{E_{elect}} + {E_{amp}}{r^2}} \right) E[ Eh|N=n ]=npk[ (E[ Nc ]+1)(2R/2mr)(2Eelect+Eamp(2r)2)+Eelect ] E\left[ {{E_h}|N = n} \right] = npk\left[ {\left( {E\left[ {{N_c}} \right] + 1} \right)\left( {2R/2mr} \right)\,\left( {2{E_{elect}} + {E_{amp}}{{\left( {2r} \right)}^2}} \right) + {E_{elect}}} \right]

From this, it can be obtained that the total energy consumption of cognitive network processing kbits data network is: E[ Et|N=n ]=E[ Eh|N=n ]+E[ Em|N=n ]=λπR2k[ (32p+2Rmr)Eelect+(1p+4Rmr)Eampr2 ] \matrix{ {E\left[ {{E_t}|N = n} \right] = E\left[ {{E_h}|N = n} \right] + E\left[ {{E_m}|N = n} \right]} \hfill \cr { = \lambda \pi {R^2}k\left[ {\left( {3 - 2p + {{2R} \over {mr}}} \right){E_{elect}} + \left( {1 - p + {{4R} \over {mr}}} \right){E_{amp}}{r^2}} \right]} \hfill \cr }

Therefore, the expectation E [D | N = n] of the total amount of data that the cognitive network can support to exchange can be calculated, which is expressed as follows: E[ D|N=n ]=nE0E[ Et|N=n ]=E0[ (32p+2Rmr)Eelect+(1p+4Rmr)Eampr2 ]k E\left[ {D|N = n} \right] = {{n{E_0}} \over {E\left[ {{E_t}|N = n} \right]}} = {{{E_0}} \over {\left[ {\left( {3 - 2p + {{2R} \over {mr}}} \right){E_{elect}} + \left( {1 - p + {{4R} \over {mr}}} \right){E_{amp}}{r^2}} \right]k}}

When the network stops working, the remaining energy (Eres)i, i = 1, 2, ⋯, n of node i is less than the energy consumption E [Eh | N = n] of the cluster head node processing unit data. Therefore, we design a double-threshold cluster head rotation algorithm for cognitive radio networks. Let Emem denote the energy consumption of any ordinary node to transmit E data to the cluster head node, which is expressed as follows: Emem=k(2Pelect+Pampr2) {E_{mem}} = k\left( {2{P_{elect}} + {P_{amp}}{r^2}} \right)

Then the total energy E0 of node i during network operation can be expressed as follows: E0=((Eh)iEmem)(Dh)i+EmemD {E_0} = \left( {{{\left( {{E_h}} \right)}_i} - {E_{mem}}} \right){\left( {{D_h}} \right)_i} + {E_{mem}}D

Among them, (Dh)i represents the number of kbits data packets that node i needs to process when it acts as the cluster head. It can be expressed as follows: (Dh)i=E0(Eres)iEmemDEhEmem {\left( {{D_h}} \right)_i} = {{{E_0} - {{\left( {{E_{res}}} \right)}_i} - {E_{mem}}D} \over {{E_h} - {E_{mem}}}}

Therefore, the cluster head energy when the second energy threshold is triggered is: (Eth2)i=(Eres)i(Eh)i(Dh)i=(Eres)i(E0(Eh)iEmemD)1[ Emem/(Eh)i ] {\left( {{E_{th2}}} \right)_i} = {\left( {{E_{res}}} \right)_i} - {\left( {{E_h}} \right)_i}{\left( {{D_h}} \right)_i} = {\left( {{E_{res}}} \right)_i} - {{\left( {{E_0} - {{\left( {{E_h}} \right)}_i} - {E_{mem}}D} \right)} \over {1 - \left[ {{E_{mem}}/{{\left( {{E_h}} \right)}_i}} \right]}}

Since the cognitive nodes are not uniformly distributed in the circular area, the number of common nodes managed by each cluster head is different, and the data forwarded by them will also be different depending on their location. The cluster head node still needs to process Dh pieces of data before the next rotation, and these data can be subdivided into data that must be processed and data that can be optionally processed. As shown in the following formula. Dh=Dh_nec+Dh_opt {D_h} = {D_{h\_nec}} + {D_{h\_opt}}

Dh_nec represents the number of data that the cluster head must process. It includes the number of data transmitted to the cluster head when the nodes in the cluster communicate, the number of data sent by the cluster head to ordinary nodes and itself, and the number of data when the cluster head detects the communication of the main user. Dh_opt represents the number of data forwarded by the cluster head to process other clusters through this cluster head. Let Ec_nec denote the energy consumption that the cluster head has to process the data. Its specific expression is as follows: Ec_nec=k(E[ Nc ]+1)[ Pelect+(2Pelect+Pamp(2r)2) ]=kp(3Pelect+4Pampr2) {E_{c\_nec}} = k\left( {E\left[ {{N_c}} \right] + 1} \right)\left[ {{P_{elect}} + \left( {2{P_{elect}} + {P_{amp}}{{\left( {2r} \right)}^2}} \right)} \right] = {k \over p}\left( {3{P_{elect}} + 4{P_{amp}}{r^2}} \right)

Therefore, when the first energy threshold is triggered, the energy consumption of the cluster head node to process the optional processing data is not greater than Eth1: (Eth1)i=(EhnpEc_nec)(Dh)i=kmrp[ (2R+mrp3mr)Eelect+(Rmr)4Eampr2 ] {\left( {{E_{th1}}} \right)_i} = \left( {{{{E_h}} \over {np}} - {E_{c\_nec}}} \right){\left( {{D_h}} \right)_i} = {k \over {mrp}}\left[ {\left( {2R + mrp - 3mr} \right){E_{elect}} + \left( {R - mr} \right)4{E_{amp}}{r^2}} \right]

After the cognitive node reaches the threshold Eth1, it will refuse to relay the data of other clusters and only serve the nodes in this cluster. When waiting for the cluster head rotation message or when the threshold Eth2 is reached, the cluster head rotation operation of the whole network is initiated.

Simulation results

M1=six and M2=3 are used in the simulation. Here, the value of N is generally related to the node density. The K value is generally related to the business mode of the node. The adjustment of these two parameters can limit the time ratio between the random access time and the scheduled network access period, and then directly affect and adjust the delay and energy consumption. In this way, various types of service requirements are met and the QoS function is improved. In the experiment, the characteristics of balanced media access delay, sleep time ratio, and balanced packet connection success rate were mainly tested.

The multimedia access delay of the message is the difference between the time when the message is correctly accessed and the time when the MAC layer receives the message delivered by the upper layer, s. The sleep time ratio refers to the total number of time slots that the node maintains the SL state in all time slots. The average packet reception success rate is the number of correctly received packets among all the normally received packets. The packet is counted as a normally received packet only when each neighbor node receives the packet normally.

The main purpose of the simulation is to test the characteristics of the SSMAC protocol. Because all nodes in the network system are in a state of having service packets to be transmitted. It is assumed that the number of service packets generated by the node conforms to the exponential distribution, so the balanced arrival time interval of the packets is 0.4~2.5s. The MAC protocol generally focuses on the data packet transmission of the local node. When a node transmits a message, it will randomly select a destination address from its neighbors as the next hop.

The protocol designed in this paper is completed through the OPNET simulation platform. It is assumed that the wireless communication module at the physical layer uses the network controller chip TR1000 of RFM Company. The chip is a low-to-medium-range, low-power and low-data-rate low-level module for wireless sensor network applications. Its average power during transmit, receive and sleep is 24.75, 13.5 and 15 μW, respectively. The maximum state change time is 20 μs. The fifty nodes in the network are uniformly distributed in the × test environment. And the average propagation 0.5 path of each node is 100m. Thus, the average number of neighbors per node is six. The average number of two-hop neighbors is seventeen. The average simulation time is 400s. The characteristics of the protocol in unicast and broadcast modes can be obtained through multiple simulations. 5~7 in the figure are the simulation result graphs of the success rate of the message, the proportion of sleep time and the access delay of multimedia equipment respectively.

Figure 5

Simulation result of message reception success rate

Figure 6

The simulation result of the proportion of message sleep time

Figure 7

Simulation results of media access delay

The simulation results show that the success rate of SSMAC packet reception has been kept below 90%. TRAMA has a very high connection success rate only in the case of light load. For the random connection of SSMAC, the overhead of the competition sector is larger than that of TRAMA at light load, so TRAMA has better energy saving than SSMAC at light load. The percentage of time occupied by the random connection period and the scheduled connection period of SSMAC is generally adjusted by two basic parameters, N and K. When N and K are fixed, the power consumption of the SSMAC agreement is basically the same, but the total power of the SSMAC agreement is basically the same under light and heavy loads. Because both SSMAC and TRAMA agreements are agreements issued according to scheduling. The success rate of traditional packet collection of competitors is lower than that of the new agreement using scheduling. However, the transmission delay of the new protocol using contention is relatively short. However, the end-to-end delay of packet transmission in the CSMA protocol is generally in the order of magnitude of 10-2 s. However, the end-to-end delay of multimedia access in SSMAC and TRAMA is generally larger than that in the competition agreement. However, the high packet reception efficiency based on the scheduling protocol can also reduce the retransmission of higher-layer packets. The reduction of Chongqing media also means that energy consumption can be reduced. Therefore, the increased end-to-end delay brought by the scheduling algorithm can be accepted in general business environments. The end-to-end delay of multimedia access of SSMAC protocol has great advantages under heavy load. Moreover, the delay of TRAMA is acceptable when the network load is light. In addition, the SSMAC protocol also stipulates that the data packets must meet the corresponding number before participating in the competition agreement. Moreover, when the parameter K is specified, when the network load is light, the packets need to wait longer in the node cache to obtain the total amount of packets participating in the agreement. This will increase the delay. In unicast mode, exposed terminals can also join the competition. Exposing ports can also improve the parallel transmission capability of the network. In the broadcast mode, since the adjacent nodes are all destination addresses, the exposed ports cannot join the competition. This will greatly affect the ability of the handshake competition between the two parts. Therefore, the parallel transmission speed of the entire Internet will be reduced, and the Internet throughput efficiency will also be lower than the traditional unicast mode.

The SSMAC protocol is a MAC protocol designed for wireless sensor networks. It not only supports the traditional peer-to-peer transmission mode, but also supports the management mode of centralized transmission. This modeling can be abstractly defined as a schema of a data aggregation tree. The sink node is regarded as the root node, while other nodes are regarded as the central node of the tree (Figure 8). The sink node is the only data aggregation node, and the data of other sensor nodes all take the sink node as the destination address. It adopts a multi-hop forwarding method, and periodically gathers the sensed data to this root node. Assuming that there is only one sink node in the simulation environment, the data of other sensor nodes are sent to the sink node hop by hop through a simple shortest path routing method. All business data of sensor nodes are generated periodically. A large data packet is generated every specified time. The simulation time still uses 400s. We can consider the successful access probability and average delay of the SSMAC protocol. Figures 9 and 10 are schematic diagrams of the simulation results of the packet access success rate and the media access delay, respectively.

Figure 8

Sink Communication Mode

Figure 9

Simulation result of message reception success rate in sink mode

Figure 10

Simulation result of media access delay in sink mode

The simulation results also show that the SSMAC protocol in the aggregate transmission mode still maintains the characteristic that the end-to-end delay of data access is relatively reduced when the load is heavier. Moreover, the SSMAC protocol can also ensure a greater success rate of packet access. And the end-to-end delay of multimedia connection when TRAMA agreement is in the aggregate transmission mode is relatively less. Moreover, it can quickly forward data packets to the sinK node when reducing the load. Therefore, by adjusting the basic parameter K of the SSMAC protocol, the end-to-end delay of the packet can be significantly reduced. And these basic parameters also determine the time judgment threshold for the SSMAC agreement to start the reservation process. The smaller the value of K, the smaller the ratio of the entire time slot occupied in the scheduling access cycle. The more frequent the reservation process is opened by the SSMAC protocol, the smaller the end-to-end delay of the packet. Because the SSMAC protocol can achieve a smaller or the same packet delay than the TRAMA protocol when the value of K is small. The experimental results show that the SSMAC protocol can be adapted to work in the convergent transmission mode of wireless sensor networks.

Conclusion

This paper mainly studies the MAC technology in wireless sensor network system. On the basis of analyzing the existing MAC layer technology, this paper presents a new MAC technology-SSMAC technology in wireless sensor network system based on scheduling technology. The technique incorporates energy-limiting and reservation scheduling techniques in WSNs. On the other hand, it fully considers the energy-saving characteristics of the network system, and has less multimedia access delay. The packet access success rate is basically the same under different loads. On the other hand, it overcomes the problems of hidden ports and exposed terminals. It realizes the QoS guarantee capability of differentiating services on the data link layer. Under the configuration of similar physical layer and simulation, the performance of SSMAC technology is also improved compared with TRAMA technology which adopts the same scheduling technology. Especially in the case of heavy load, it can bring more stable packet access success rate and sleep time ratio, and less network access delay.

Figure 1

Time frame diagram
Time frame diagram

Figure 2

Schematic diagram of reservation time slot
Schematic diagram of reservation time slot

Figure 3

Contention message format
Contention message format

Figure 4

The relationship between network energy consumption and p and λ per unit time
The relationship between network energy consumption and p and λ per unit time

Figure 5

Simulation result of message reception success rate
Simulation result of message reception success rate

Figure 6

The simulation result of the proportion of message sleep time
The simulation result of the proportion of message sleep time

Figure 7

Simulation results of media access delay
Simulation results of media access delay

Figure 8

Sink Communication Mode
Sink Communication Mode

Figure 9

Simulation result of message reception success rate in sink mode
Simulation result of message reception success rate in sink mode

Figure 10

Simulation result of media access delay in sink mode
Simulation result of media access delay in sink mode

Costes, J. M., Kairouz, S., Monson, E., & Eroukmanoff, V., (2018). Where lies the harm in lottery gambling? A portrait of gambling practices and associated problems. Journal of gambling studies, 34(4), 1293–1311. CostesJ. M. KairouzS. MonsonE. EroukmanoffV. 2018 Where lies the harm in lottery gambling? A portrait of gambling practices and associated problems Journal of gambling studies 34 4 1293 1311 10.1007/s10899-018-9761-329536292 Search in Google Scholar

Hosahalli, D., & G. Srinivas, K., (2020). Cross-layer routing protocol for event-driven M2M communication in IoT-assisted smart city planning and management: CWSN-eSCPM. IET Wireless Sensor Systems, 10(1), 1–12. HosahalliD. G. SrinivasK. 2020 Cross-layer routing protocol for event-driven M2M communication in IoT-assisted smart city planning and management: CWSN-eSCPM IET Wireless Sensor Systems 10 1 1 12 10.1049/iet-wss.2018.5198 Search in Google Scholar

Dwivedi, A. K., Sharma, A. K., & Mehra, P. S., (2020). Energy efficient sensor node deployment scheme for two stage routing protocol of wireless sensor networks assisted iot. ECTI Transactions on Electrical Engineering, Electronics, and Communications, 18(2), 158–169. DwivediA. K. SharmaA. K. MehraP. S. 2020 Energy efficient sensor node deployment scheme for two stage routing protocol of wireless sensor networks assisted iot ECTI Transactions on Electrical Engineering, Electronics, and Communications 18 2 158 169 10.37936/ecti-eec.2020182.240541 Search in Google Scholar

Zhu, K., Hu, J., Liu, Y., Alotaibi, N. D., & Alsaadi, F. E., (2021). On ℓ2–ℓ∞ output-feedback control scheduled by stochastic communication protocol for two-dimensional switched systems. International Journal of Systems Science, 52(14), 2961–2976. ZhuK. HuJ. LiuY. AlotaibiN. D. AlsaadiF. E. 2021 On ℓ2–ℓ∞ output-feedback control scheduled by stochastic communication protocol for two-dimensional switched systems International Journal of Systems Science 52 14 2961 2976 10.1080/00207721.2021.1914768 Search in Google Scholar

Rawat, P., & Chauhan, S., (2021). Particle swarm optimization-based energy efficient clustering protocol in wireless sensor network. Neural Computing and Applications, 33(21), 14147–14165. RawatP. ChauhanS. 2021 Particle swarm optimization-based energy efficient clustering protocol in wireless sensor network Neural Computing and Applications 33 21 14147 14165 10.1007/s00521-021-06059-7 Search in Google Scholar

Ma, H., Teng, J., Hu, T., Shi, P., & Wang, S., (2020). Co-communication protocol of underwater sensor networks with quantum and acoustic communication capabilities. Wireless Personal Communications, 113(1), 337–347. MaH. TengJ. HuT. ShiP. WangS. 2020 Co-communication protocol of underwater sensor networks with quantum and acoustic communication capabilities Wireless Personal Communications 113 1 337 347 10.1007/s11277-020-07192-7 Search in Google Scholar

Wang, Z. X., Zhang, M., Gao, X., Wang, W., & Li, X., (2019). A clustering WSN routing protocol based on node energy and multipath. Cluster Computing, 22(3), 5811–5823. WangZ. X. ZhangM. GaoX. WangW. LiX. 2019 A clustering WSN routing protocol based on node energy and multipath Cluster Computing 22 3 5811 5823 10.1007/s10586-017-1550-8 Search in Google Scholar

Zou, L., Wang, Z., Hu, J., Liu, Y., & Liu, X., (2021). Communication-protocol-based analysis and synthesis of networked systems: progress, prospects and challenges. International Journal of Systems Science, 52(14), 3013–3034. ZouL. WangZ. HuJ. LiuY. LiuX. 2021 Communication-protocol-based analysis and synthesis of networked systems: progress, prospects and challenges International Journal of Systems Science 52 14 3013 3034 10.1080/00207721.2021.1917721 Search in Google Scholar

Souza, G. C., Giacomin, J. C., & Heimfarth, T., (2021). An Asynchronous Anycast Protocol Resilient to Changes in Communication Channel. Wireless Personal Communications, 120(4), 3243–3263. SouzaG. C. GiacominJ. C. HeimfarthT. 2021 An Asynchronous Anycast Protocol Resilient to Changes in Communication Channel Wireless Personal Communications 120 4 3243 3263 10.1007/s11277-021-08611-z Search in Google Scholar

Sai, M. S. S., Rejeti, I. D. V. K. K., Gera, I. D. J., & Raju, M. N., (2021). EFFECTIVE ROUTING PROTOCOL IN MOBILE AD-HOC NETWORK USING INDIVIDUAL NODE ENERGY. Technology (IJARET), 12(2), 445–453. SaiM. S. S. RejetiI. D. V. K. K. GeraI. D. J. RajuM. N. 2021 EFFECTIVE ROUTING PROTOCOL IN MOBILE AD-HOC NETWORK USING INDIVIDUAL NODE ENERGY Technology (IJARET) 12 2 445 453 Search in Google Scholar

Gomathy, V., Padhy, N., Samanta, D., Sivaram, M., Jain, V., & Amiri, I. S., (2020). Malicious node detection using heterogeneous cluster based secure routing protocol (HCBS) in wireless adhoc sensor networks. Journal of Ambient Intelligence and Humanized Computing, 11(11), 4995–5001. GomathyV. PadhyN. SamantaD. SivaramM. JainV. AmiriI. S. 2020 Malicious node detection using heterogeneous cluster based secure routing protocol (HCBS) in wireless adhoc sensor networks Journal of Ambient Intelligence and Humanized Computing 11 11 4995 5001 10.1007/s12652-020-01797-3 Search in Google Scholar

Sai, M. S. S., Rejeti, I. D. V. K. K., Gera, I. D. J., & Raju, M. N., (2021). EFFECTIVE ROUTING PROTOCOL IN MOBILE AD-HOC NETWORK USING INDIVIDUAL NODE ENERGY. Technology (IJARET), 12(2), 445–453. SaiM. S. S. RejetiI. D. V. K. K. GeraI. D. J. RajuM. N. 2021 EFFECTIVE ROUTING PROTOCOL IN MOBILE AD-HOC NETWORK USING INDIVIDUAL NODE ENERGY Technology (IJARET) 12 2 445 453 Search in Google Scholar

Gomathi, S., & Gopala Krishnan, C., (2020). Malicious node detection in wireless sensor networks using an efficient secure data aggregation protocol. Wireless Personal Communications, 113(4), 1775–1790. GomathiS. Gopala KrishnanC. 2020 Malicious node detection in wireless sensor networks using an efficient secure data aggregation protocol Wireless Personal Communications 113 4 1775 1790 10.1007/s11277-020-07291-5 Search in Google Scholar

Sai, M. S. S., Rejeti, I. D. V. K. K., Gera, I. D. J., & Raju, M. N., (2021). EFFECTIVE ROUTING PROTOCOL IN MOBILE AD-HOC NETWORK USING INDIVIDUAL NODE ENERGY. Technology (IJARET), 12(2), 445–453. SaiM. S. S. RejetiI. D. V. K. K. GeraI. D. J. RajuM. N. 2021 EFFECTIVE ROUTING PROTOCOL IN MOBILE AD-HOC NETWORK USING INDIVIDUAL NODE ENERGY Technology (IJARET) 12 2 445 453 Search in Google Scholar

Velusamy, D., Pugalendhi, G., & Ramasamy, K., (2019). A cross-layer trust evaluation protocol for secured routing in communication network of smart grid. IEEE Journal on Selected Areas in Communications, 38(1), 193–204. VelusamyD. PugalendhiG. RamasamyK. 2019 A cross-layer trust evaluation protocol for secured routing in communication network of smart grid IEEE Journal on Selected Areas in Communications 38 1 193 204 10.1109/JSAC.2019.2952035 Search in Google Scholar

Chen, S., Guo, J., & Ma, L., (2019). Sliding mode observer design for discrete nonlinear time-delay systems with stochastic communication protocol. International Journal of Control, Automation and Systems, 17(7), 1666–1676. ChenS. GuoJ. MaL. 2019 Sliding mode observer design for discrete nonlinear time-delay systems with stochastic communication protocol International Journal of Control, Automation and Systems 17 7 1666 1676 10.1007/s12555-018-0727-0 Search in Google Scholar

Articles recommandés par Trend MD

Planifiez votre conférence à distance avec Sciendo