Energy-Efficient Duty-Cycle Hybrid Medium Access Control Protocol for Wireless Sensor Network

This paper proposes an energy-efficient hybrid medium access control protocol based on precise control of duty cycle. To save energy, the proposed protocol optimally adjusts the trans-receiver radio's turn-on and off time. By estimating slot utilization, both the cluster head and the sensing devices save energy. During the reserved slot duration, each sensor device periodically checks its buffer status and turns off its radio when the buffer becomes empty. Each sensor device sends a slot utilisation status to the cluster head device. The cluster head device also turns its radio on and off based on the slot utilisation status of sensing devices. For optimal energy consumption, a duty cycle approach is proposed. According to the mathematical and simulation results, the proposed protocol consumes less energy than existing equivalent protocols. Simulation results depict, exactly similar results of analytical and mathematical analysis prove the validity of the work.


Introduction
For numerous monitoring applications, wireless sensor networks are frequently employed [1][2][3]. Researchers have proposed a few application-specific and conventional medium access control (MAC) protocols for various applications. Most general monitoring applications can use a standard IEEE 802.15.4 MAC protocol [4]. The superframe architecture of the standard IEEE 802.15.4 MAC protocol can be used for a variety of monitoring applications, depending on the requirements. Both contention and reservation-based transmission can be utilized in standard IEEE 802. 15.4 MAC protocol. In the contention access period (CAP), carrier sense multiple access/collision avoidance (CSMA/CA) protocol is used for the transmission of data. Reservation-based data transmission slots are utilized for data transmission during a contention-free period (CFP). However, due to its inefficient use of the contention period, the performance of the standard IEEE 802.15.4 MAC protocol is not acceptable for many protocols. The researchers have proposed different application-specific protocols for the specific application needs [4]. For railway monitoring applications, G.M. Shaifullah developed an energy-efficient bit-map assisted (EBMA) MAC protocol [5]. The sensing devices' energy usage is reduced by using the piggybacking method to reserve the next slot. For incident-focused and constant-checking hybrid applications, Tolani et al. suggested an energy-efficient hybrid MAC (EEHMAC) protocol [6]. EA-TDMA [7] and BMA MAC [8] capabilities are included in the protocol. In EA-TDMA slots, the protocol may handle data packet transmission with a time constraint. However, with the EEHMAC protocol, energy consumption can be lowered by optimizing the trans-receiver radio's on-off period. In this paper, we have proposed the energy-efficient duty cycle hybrid MAC (EEDCHMAC) protocol to improve the protocol's efficiency. Each session is separated into two sub-slots in the proposed protocol. The sensor device transmits data using the EA-TDMA MAC protocol in sub-slot-1. Sensing devices send data using the dutycycle-bit-map-assisted (DC-BMA) MAC protocol in sub-slot-2. The proposed scheme efficiently turn-on and off the trans-receiver device. Each sensor device continuously monitors its buffer in reserved slot duration. When the buffer becomes empty, the sensor device turns off its radio. Therefore, the sensing devices save energy to the devices which partially utilize the reserved slot. To save the energy consumption of the cluster head device, a duty cycle arrangement is done in the proposed scheme. Each sensor device transmits its buffer status to the cluster head device. For the transmission of buffer status, 2 bits of the bit-mapping period are utilized. In the proposed protocol both the cluster head device and sensor device efficiently turn on and off the trans-receiver radio to efficiently control the energy consumption.
The rest of the paper's structure is explained below. Related recent works are discussed in Sect. 2. The EEDCHMAC explanation is described in Sect. 3. The mathematical modeling portion is described in Sect. 4. The simulation analysis and results are presented in Sect. 5. The conclusions are summarized in Sect. 6.

Related Works
The researchers have reported various works deals with the energy-efficient contention protocols. The recent contributions of the researchers are explained below in Table 1.
In [9], the Quintero et al. proposed energy efficient MAC protocol for energy harvesting devices. In [10], Jimenez et al. proposed duty cycle MAC protocol. In [11], Aljabi et al. proposed hybrid MAC protocol for dynamic adaption of sleep/wakeup time. In [12], MAC protocol is proposed for the cognitive radio application. In [13], the author proposed QoS duty cycle MAC protocol. Similarly in [14][15][16][17], the authors have proposed different methods of the energy-efficient data transmission. However, in any of the method the researchers have not proposed efficient switching of the trans-receiver radio for the precise duty cycle control. Therefore, we have proposed energy-efficient hybrid duty cycle MAC protocol.  [11] To develop a hybrid protocols to dynamically adapt sleep/ awake time for dynamic load application To utilize the nodes' energy and dynamically adapt the sleep/ wake-up times towards the network demands, a hybrid time TDMA-CSMA/CA MAC protocol was created Syed et al. [12] To reduce the energy consumption of battery-driven devices for cognitive radio application Group control slot allocation (GCSA) protocol, an energy-efficient medium access technique, is used in the current study for cognitive radio (CR) networks Muzzakari et al. [13] To provide high QoS for priority aware data transmission Reduce the time it takes for high priority packets to arrive by using an Energy Efficient and QoSaware (EEQ) MAC protocol with a duty cycle scheme that adjusts the node's duty cycle to the queue size and priority class of a packet Kim et al. [14] To reduce the number of redundant convolution layers, maxpooling operations are used A DCNN accelerator with a novel conditional computing scheme that synergistically combines precision cascading (PC) and zero skipping is proposed (ZS) Precision cascading is proposed, in which the input features are divided into several low-precision groups and approximate convolutions are used Figure 1 depicts the EEDCHMAC protocol's operational diagram. The proposed MAC protocol's operation begins with the setup step. The proposed MAC protocol's setup phase is comparable to that of previously described protocols such as TDMA [18], EA-TDMA [7], BMA, E-BMA [5], ABMA [19], ASH-MAC [20], and EEHMAC [6]. The EEDCH-MAC protocol's post-setup phase is similar to that of the ASH-MAC and EEHMAC protocols. Sensing devices are divided into two categories during the post-setup phase: constant-checking and incident-focused devices. As illustrated in Fig. 1, the data transmission phase began following the post-setup phase. Each session is split into two sub-slots during the data transmission phase. Similar to the ASH-MAC and EEHMAC protocols, constantchecking devices send data during sub-slot-1 while incident-focused devices transmit data during sub-slot-2. The EA-TDMA MAC protocol, which is comparable to the EEHMAC protocol, is used by sub-slot-1. The sub-slot-2 procedure, on the other hand, is distinct from the existing MAC protocols. Sub-slot-2 of the proposed EEDCHMAC protocol uses a duty cycle method with a bit-map-assisted MAC protocol, known as DC-BMA. Each source device broadcasts a two-bit buffer status to the cluster head device via the DC-BMA MAC protocol. During the whole contention period, all non-source devices are in a sleep state. Based on the filled buffer, the source devices broadcast a 2-bit status. Table 2 shows the bit status of the filled buffer. All of the sensing devices keep an eye on their buffer and turn off their radio when it becomes too full. However, the cluster head device switches off its radio based on the buffer status obtained from the sensing devices. The precise control of trans-receiver radio turn-on and off times saves the overall energy consumption. Increasing the number of bits of the buffer status reduces the quantization error. Hence, controls the trans-receiver radio more optimally. However, it increases the transmission bandwidth of the control packets. In the proposed protocol, we have used 2-bit status which can be increased as per the application need.

Analytical Model for Energy Consumption (EEDCHMAC)
The cluster head allots one reserved slot to all sensing devices that require it during the contention phase, according to the conventional rule of bit-map-assisted media access control protocol. Many of the devices, on the other hand, only use a fraction of the authorized slot. Assume that each device makes use of the fractional portion of the slot. Let's say there are total N devices, and out of those N sensing devices, m is constant-checking devices and N-m sensing devices are incident-focused sensing devices. The data is transmitted in subslot-1 by constant-checking devices using the EA-TDMA MAC protocol [7], and in subslot-2 by incident-focused devices using the DC-BMA MAC protocol [8]. Therefore, we can assume that the utilization factors of the N-m sensing devices are µ 1 , µ 2, µ 3, … µ N-m . The average utilization of N-m incident-focused sensing devices is given by, Each sensor device continuously monitors its buffer status and turns off its radio when the buffer becomes empty, according to the offered approach. As a result, we can use Eq. 1 to determine the entire energy consumption of SNs.
The EEDCHMAC has a post-setup step, similar to the ASH-MAC [20] and EEHMAC [6], to classify the SNs into constant-checking and event-monitoring devices. ASH-MAC [20] has already calculated the energy consumption for post-setup phase as given below: The sub-slot-1 of the EEDCHMAC protocol is comparable to the EEHMAC protocol [6]. As a result, the total energy consumption in sub-slot-1 can be calculated as follows: Refer to [6] for the derivation. The slot utilization factors of sensing devices during sub-slot-1 are U1, U2, and U3. The EEDCHMAC protocol's sub-slot-2 differs from the EEHMAC protocol, as previously stated. We used a duty cycle-based bit-map-assisted MAC algorithm in the EEDCHMAC protocol. As a result, the energy consumption of the sensor device and cluster head device is calculated in this section.
Sub-slot-2 starts with a contention phase identical to the BMA, which is repeated for each Sub-slot-2. All constant-checking devices turn off their radios during Sub-slot-2. Because there are (N-m) incident-focused devices, the contention phase has (N-m) control slots. During the Sub-slot-2 phase, all incident-focused devices turn on their radios. Each incident-focused device sends a control packet within their given slot if there is data to  N-m-1) slots, the source devices are idle. The following is the energy consumption of sub-slot-2: The data transmission phase occurs after the contention phase, and each source device sends the data packet in its assigned data slot. Non-source devices turn off their radios in the same way that BMA does during Sub-slot-2. On the other side, the offered technique employs duty cycle-based BMA. Each sensor device continuously checks its buffer and turns off its radio when the buffer is empty. The energy consumption of a sensor device during the transmission phase of Sub-slot-2 is calculated as follows: Equation 1 shows how to calculate the value of the overall utilization factor. The buffer state of each SN is sent to the CH device. The CH device turns on and off its radio based on the filled buffer condition. During the contention phase, each sensor device broadcasts a 2-bit buffer status. Two-bits are used in four different levels. Therefore, we can assume that the quantized value of utilization factor of the N-m sensing devices is µ 1 q , µ 2 q , µ 3 q , … µ N-m q . The average quantized utilization of N-m incident-focused sensing devices is given by, The overall energy consumption of cluster head for N-m slots can be given by, Therefore, the total energy consumption for the transmission phase of sub-slot-2 can be given by, The total energy consumption of the EEDCHMAC protocol in k sessions can be given by,

Result Analysis and Discussion
The simulation analysis can be done on the simulation background which is used in the previously reported works. This will help to verify the validity of the result. To ensure the validity of the results, we have compared the analytical and simulation results. As the event-based data traffic is unpredictable and uncertain. Therefore, we have considered three different data traffic load conditions i.e. low, medium, and high data traffic. For analysis of low data traffic load condition, we have assumed that for low data traffic condition, only up to 25% slot is utilized by the incident-focused devices. For medium and high data traffic conditions, it is assumed the respectively 50% and 100% reserved data slot is utilized by the sensing devices. For these three data traffic conditions, we have analyzed the network for four different scenarios. In the first scenario, the event-occurrence probability of the source device is varied. In the second scenario, the number of sensing devices is varied i.e. network size is increased. In the third and fourth scenarios, data packet size and number of rounds are varied respectively. For simulation and analytical investigation, the ZigBee-enabled 2.4 GHz CC2420 RF transceiver is used [4]. The transmitter, receiver, and idle state power consumption are taken as per the ratings of the CC2420 RF module.

Scenario-1
In the first scenario, we have analyzed the network for the probability of occurrence of the event. For the non-source devices, the event occurrence probability will be zero. As the event probability increases, the data generation rates increases, and therefore the number of source devices increases. This result shows the higher energy consumption of the sensing devices and also the cluster head device.
As shown in Fig. 2, the energy consumption linearly increases with probability. However, the energy consumption of the offered protocol is lower w.r.t. ASH-MAC and EEHMAC protocol in all three conditions. Even at high data traffic load conditions, the offered EEDCHMAC protocol saves a satisfactory amount of energy. This is because the offered protocol continuously monitors the buffer and turn-off its radio in idle condition when the buffer becomes empty. For the acknowledgment of the cluster head device, the offered protocol follows the duty cycle approach and each sensor device transmits its buffer status. Based on the buffer status of the sensing devices, the cluster head device precisely controls the turn-on and off times of the radio. The cross-layer effort from the MAC layer to the physical layer improves the performance of the network in terms of energy consumption.

Scenario-2
In this scenario, the network is analyzed for network size. The network size is increased by increasing the number of devices. As the network size increases, the network load increases, and energy consumption also increases. However, the energy consumption of the offered protocol even in the worst condition is much better than the other existing MAC protocols. The energy consumption of the offered MAC protocol is 20-25% less w.r.t. the ASH-MAC and 15-20% less than EEHMAC protocol as shown in Fig. 3. The EEDCH-MAC protocol saves energy by precise control of turn-on and off duration of trans-receiver radio. A detailed discussion has already been done in scenario-2.

Scenario-3
In this scenario, the network is analyzed for the different data packet sizes. As the data packet size increases, the utilization of the channel improves as shown in Fig. 4. For the larger data packet size, the devices more precisely control the trans-receiver radio, and at low data traffic load conditions, the non-source devices increase. Therefore, the energy consumption of the network at low data traffic conditions is much lower than the existing protocol. Even in the worse condition at high traffic load, the energy consumption of EED-CHMAC is lower than the EEHMAC and ASH-MAC protocols.

Scenario-4
In this scenario, the network is analyzed for the increased number of rounds in each session.
As the number of rounds increases, the overall energy consumption increases as shown in Fig. 5. The energy consumption of the offered EEDCHMAC protocol is much lower than the existing MAC protocols. The offered protocol controls the trans-receiver radio more precisely. Therefore, the energy consumption of the offered MAC protocol is lower than the ASH-MAC and EEHMAC protocols.
For the better representation of the results, the comparison of maximum energy consumption with respect to probability of packet generation, number of nodes, packet size, and number of rounds are given in Table 3. The statistics also proves the superiority of the proposed protocol.

Conclusion
In this paper we have proposed an energy-efficient duty cycle MAC protocol. The results prove that each sensing device efficiently controls the trans-receiver radio in the proposed protocol. Each sensor device continuously monitors its buffer status and turns off its radio when the buffer becomes empty. As a result, the network's overall energy consumption decreases. The results show that the proposed protocol saves 30% more energy with respect to EEHMAC and 24% more energy w.r.t. ASHMAC for high data traffic (in worst case). The proposed protocol saves much more energy for moderate and low data traffic applications (for moderate data traffic the energy saving is 46% and 40% respectively. Similarly, for low data traffic energy saving is 59% and 53% respectively). The simulation results confirm the validity of the work by replicating the analytical finding.