Improving the performance of WPAN …………M.H. Improving the performance of WPAN during remote E-Health monitoring in crowded public patients room

Wireless Personal Area Network (WPAN) represents the first wireless network contact with patients in the applications of remote E-health monitoring using Wireless Sensor Network (WSN). In this paper,a way to improve the performance of the WPAN in the crowded public patient room was presentedusing a multiplied decreasing for both of packet size and packet inter-arrival time, as well as a multiple coordinating to achieve the best results in throughput and delay.


Introduction
Wireless Personal Area Networks (WPANs) are an importantnetwork type that has much potential of playing a main role in remote E-Health monitoring using Wireless Sensor Network expressed as an emerging technology in the healthcare system.This WPAN technology permits short-distance communication and servesa specialized purpose such as networking medical sensing devices in healthcare system. Patients can carry wearable sensors as Wireless Body Area Network (WBAN) that candetect vital signs (such as, glucose level, heart rate, etc.) andcan monitor the patient continuously using this WPAN [1].After that, a personal gateway nodecollects all the sensor data from the WPAN and forwardsthem to a doctor or a remote online server in the hospital. IEEE 802.15.4/Zigbee is a standard base network protocol that iswidely used for Low Rate WPAN (LR-WPAN). Zigbee and IEEE 802.15.4 are not the same. The ZigBee alliance was formed prior to the formation of the IEEE 802.15.4 group. Later, the ZigBee Alliance and the IEEE 802.15.4 group have decided to join forces and use ZigBee as the commercial name for this technology [5].

Related Work
Nahla A. and Fuad A. [7] tested the topological characteristics and performance of the IEEE802.15.4 standard in terms of throughput andnode to node delay using OPNET simulator, theyinvestigated the self-healing feature on a mesh topology from earlier design until it supports a large number of end points (>64,000) with dynamic routing.
Hammoodi I., et al. [3] Studied and analyzed the QoS performance evaluation of the ZigBee protocol within the OPNET simulator for different WSN topologies and routing schemes.
Mumtaz M. Ali Al-Mukhtar and Teeb Hussein Hadi [6] investigated the performance of WSN based on various topologies, and adoptedsome techniques useful for managing wireless sensor networks.
Ramyah S. [9] analyzed the variations in load metric in hexagonal configuration by enabling and disabling acknowledgment message (ACK) depending on IEEE 802.15.4/ZigBee standard for WSN.

Physical Layer
Zigbee uses three frequency bands for transmission of data: -868 MHz band with a single channel having a data rate of 20 kb/s. -The 915MHz band with 10 channels, and each channel has a central frequency separated from the adjacent band by 2 MHz and data rate of 40 kb/s. BPSK modulation technique is used in which symbols are transmitted at 1 bit per symbol.

MAC layer
The transmission Channel is accessed primarily through Carrier Sense Multiple Access-Collision Avoidance (CSMA-CA) protocol. The MAC layer can take care of transmitting data. The MAC layer decides whether to use slotted or unslotted CSMA-CA. It also takes care of scanning the channel, starting PANs, detecting and resolving PAN ID conflicts, and performing device discovery etc. [4].

Zigbee Network Topologies: Star Topology
In star topology, a coordinator is the prime node and all other devices are directly connected to it. Every data exchange between 2 end devices must pass through the coordinator first. This topology is very much vulnerable to collapses since the whole network goes down if the prime node fails. Employing routers are a waste of energy here as their functionality is never actually used.

Tree Topology:
In tree topology, the prime node would be the root node of the network with hierarchical body. There is a point to point connection between any 2 nodes i.e. a single path exists for reaching a node. Due to the self-healing capability of dynamic routing employed, in case of collapses, the backup would be prepared from the vicinity instantly (if available) [4].

Mesh Topology:
In mesh topology, data packets can be directly relayed between the routers and then to nodes. They need not pass through the prime node. Such a network has multiple paths for reaching a node and hence a backup can be made easily in failure situations e.g. if a router stops working then any nearby router will tackle the traffic of that router in a very finite time without affecting the performance much [11].
Star topology is commonly used in Remote Patient Monitoring. A star-based ZigBee network uses the master (coordinator) and slaver (end devices) mode. The master node is usually put in the center of a WPAN. It initiates the WPAN and controls communication within the WPAN.
There are three main types of data transmission: from a coordinator to an end device, from an end device to a coordinator and between two coordinators. The mechanisms for these transmissions depend on the mode. Figure 2 illustrates the data transmission mode between a coordinator and an end device, which is commonly used in a ZigBee WPAN. Univ. Aden J. Nat. and Appl. Sc. Vol. 22 No.1 -April 2018

THE PROPOSED ALGORITHM
We introduced while statement to improve a performance of WPAN by multiplied decreasing in both packet size (Z) and the packet inter-arrival time (T) and multiplied increasing in the number of coordinators (K). To evaluate the performance of WPAN, firstly, set the number of patients to 3, 6 and 12, and set the number of sensors to 3 (for example) for each patient.Secondly initialize the packet size to 1024 bits.Thirdly, set the inter-arrival time to 1 sec.fourthly, set the number of coordinators (K) to 1 (single coordinating), and set C to 1 that will be used as a counter for increasing the decreasing operation.Fifthly, perform the first unslottedCSMA/CA operation during the single coordinating, then evaluate the performance after applying this CSMA/CA operation. If there is a degradation in the performance, then go to applying the decreasing operation of one half for default packet size and default packet inter-arrival time.After applying the decreasing operation of one half, we will again apply the unslotted CSMA/CA operation during the single coordinating, and increase the number of coordinators by 2 for the first multiple coordinating and the next unslotted CSMA/CA operations. After that, the last evaluation in performance for the single coordinating is performed; if there is a degradation in performance, then other decreasing operation of two halves and unslotted CSMA/CA operation are performed for the first multiple coordinating for K = 2, 3 and 5 of 3, 6 and 12 patients respectively, or decreasing operation of three halves and unslotted CSMA/CA operation are performed for the second multiple coordinating for K = 3, 5 and 9 for 3, 6 and 12 patients respectively. But if there is no degradation in performance, then the WPAN is accepted. A flowchart of the algorithm that uses while statement is shown in figure 3.

SIMULATION AND CONFIGURATION
For training the algorithm, we designed a WPAN in a patient room with 3, 6 or 12 patients, and each patient has 3 sensors, each sensor is simulated by zigbee_end_device and sends its data to a destination that is simulated by zigbee_coordinator. In the single coordinating, all zigbee_end_devices send data to only one zigbee_coordinator, but in the multiple coordinating, some zigbee_end_devices send its data to the first zigbee_coordinator and other send its data to the second zigbee_coordinator and so on. Each WPAN starts at the sensors that are attached to the surface of the patient body and ends at the coordinator. If the patients are found in the same location, such as this public patient room, we can implement one WPAN for all the patients using the single coordinating, but we can implement more than one WPAN using the multiple coordinating, depending on some reasons such as the financial ability of the patient or a mobility of the patient outside hospital. Univ. Aden J. Nat. and Appl. Sc. Vol. 22 No.1 -April 2018

Scenarios
The algorithm of this first improvement level is simulated by 12 scenarios. The first four scenarios use 3 patients, the second four scenarios use 6 patients and the last four scenarios use 12 patients. The first scenarios in each group named WSN_3_3_1_nodcr, WSN_6_3_1_nodcr and WSN_12_3_1_nodcr, where 3, 6 and 12 patients are monitored respectively. Each patient has 3 sensors, all the patients are coordinated by one coordinator, and no decreasing operations in packet size and PIT are used. Only one WPAN is used in these scenarios. The second scenarios in each group named WSN_3_3_1_dcr, WSN_6_3_1_dcr and WSN_12_3_1_dcr that are like the first scenarios, but here the decreasing operations of one half (  Figures (4, 5, 6, 7, 8, 9, 10, 11 and 12).

Configuring a sensor of patient
We set specified properties when we configured the sensor (sensor1 for example) of the patient based on the Table 1.

Configuring a coordinator in WPAN
We set specified properties when we configured the coordinator (coordinator1 for example) of any WPAN based on the Table 2.

Experimental Results and Discussion
All the statistics are obtained for delay and throughput results. Figures 13 and 14 represent the statistics of the throughput and delay results respectively for 3 patients using the single coordinating (one coordinator is used) and the multiple coordinating (2 and 3 coordinators are used) with or without decreasing operations, Figures 15 and 16 also represent the statistics of the throughput and delay results respectively but for 6 patients using the single coordinating (one coordinator is used) and the multiple coordinating (3 and 5 coordinators are used) with or without decreasing operations. Figures 17 and 18 also represent the statistics of the throughput and delay results respectively but for 12 patients using the single coordinating (one coordinator is used) and the multiple coordinating (5 and 9 coordinators are used) with or without decreasing operations. From the results we got the best throughput and delay in the scenarios WSN_3_3_3_dcr, WSN_6_3_5_dcr and WSN_12_3_9_dcr where the scenarios have much more increasing in coordinators that enhanced the distribution of the single WPAN load using other WPANs, and much more increasing in the decreasing operations that increased the packetizing on the data to give small sizes of packets with the best small distance between these packets. So the multiple coordinating and the decreasing for both of packet size and packet inter-arrival time are participated in improve the performance of the patients monitoring using WPAN network technology.

Conclusions
From the results and analysis, the improvement level was implemented to enhance the performance of WPAN in public patient room using the decreasing for both the packet size and packet inter-arrival time and the multiple coordinating where an algorithm with While statement is used. We have noticed that more increasing in the throughput and more decreasing in the delay were produced when much more of both multiple coordinating and multiplied decreasing operations were performed.