Wireless sensor mote architecture
Aditya Tarey, Omkar Revankar, Rahul Singh Thakur, Vidyasagar Hatgine.
Department of Electrical Engineering, University of Texas at Dallas
February 15, 2019 1
Abstract—Environmental applications sensor network generally calls for a reliable yet energy efficient sensor network, due to the fact that it should only need a sample data at the assign time. However, the accuracy of the data is important since the result is usually far in between and each data collection will be used to study, therefore this paper will study Reliable Multi-Segment Transport (RMST) protocol in the transport layer for an environmental application.
Keywords Wireless Sensor Network, RMST, Reliable Multi-Segment Transport, Transport Layer
The wireless sensor networks have left behind the wired sensor networks as they have been proved much more feasible, scalable and cost effective when compared to the dense wired system which could cause a chaos. The widespread use of wireless sensor systems and networks is due the availability of inexpensive, low powered, miniature components like the processors, sensor circuitry, amplifier, radio which are often integrated on a single chop (system on chip) or ASIC. Moreover, the idea of Internet of Things was developed in parallel to the WSN which gave a boom to the wireless sensor network’s industry. Hence, the design and analysis of the wireless sensor nodes and components must be mined with extra care so that the overall system is cost effective and has high performance. One more additional concern about the wireless sensor networks is the data manipulations and the swiftly arising issue which is the act of making autonomous decisions which is widely known as Machine Learning which plays a wide role in the data manipulation and analysis.
II. CHARACTERISTIC FEATURE OF WSNs
A WSN can be elucidated as a network of sensor nodes that can independently and cooperatively sense the physical signals which enables the interaction between the human or computer system with the surrounding environment. The area where the sensors are deployed is known as sensor field. The main function of the sensor nodes is to sense, collect data and transmit the data to the cloud with the medium of a gateway where things are analyzed at the management node and the entire information is known by the observer.
Figure 1- Block diagram showing architecture of sensor nodes .
III. ARCHITECTURE OF SENSOR NODEs
The architecture consists of the following: power managing module, a sensor, a microcontroller and a transceiver. The main function of the power management module is to provide adequate power to the sensor system to make the system reliable. A sensor is the one which converts the physical quantity into a small electrical signal which would be further amplified by the eligible amplifier for the signal to be analysed by controller. The microcontroller computes the incoming information
and sends the information to the cloud through the transceiver (RF module) which would help realise the communication physically. It is important for all the components to be miniature and cost effective so that the entire system would be reliable .
Figure 2- Functions of the power management module.
IV. AMPLIFIER DESIGN FOR SENSOR MOTE
Low data rate sensor networks are rapidly emerging and currently the design of low cost, light weight and fully integrated sensor nodes has become a challenge to the engineers. The realisation of such fully integrated system can be done by using an application specific amplifier to increase the gain of the system by analysing the input signal. There are different types of amplifiers depending upon the nature of the available input small signal so that effective amplification of the input signal takes place so the computation of the signal gives reliable results. The differential amplifier is the one which can be used to decrease the effect of temperature and noise in the system. The differential amplifier basically consists of two stages. The role of the first stage is to convert the voltage into current and then the current is mirrored into the second stage and it is converted into voltage .
Figure 3- Differential amplifier design.
V. MOTE PLATFORM FOR WIRELESS SENSOR NODES
Nowadays, the research in WSN sensor nodes has reached to a level where researchers have been using rich computing platforms that would be using the FPGA accelerators. Technological advances and the demanding applications have made the WSN industry use different types of mote platforms such as FPGA as processing unit, microcontroller, DSP, ASIC/SoC and multicore architectures. In this paper we throw light on the wireless sensor nodes based on a microcontroller.
Figure 4- Block diagram of MCU unit.
The MCU is the main processing unit of the wireless sensor networks and systems. The nodes mainly consist of the following parts: ADC, memory unit, I/O unit and the battery unit. The main function of the ADC is to covert the incoming analog signal which would the MCU to compute it.
The computing unit is the brain of the system carrying out the processing tasks of the nodes like data filtering, data aggregation, data compression, and data routing. The processor may vary from ATMEGA 128 to ARM Cortex M3.The memory unit holds the data and instruction got processing information. It contains RAM, Flash memory and EEPROM. The communication module allows the nodes interconnection and data exchange. It is generally based on Radio frequency module (RF Module). Other modules such as Acoustic, Microwave, infrared etc. can also be used based on the application .
VI. EDGE COMPUTING IN WIRELESS SENSOR NETWORK
Edge computing leads the way in bringing Memory and Computer Power close to desired location of the user. Edge computing helps in optimizing applications or cloud computing systems it takes away minor segment from an application, data, or services from one or more central nodes referred as “node” to the other logical extreme called as “edge”
Figure 5- Block diagram of main flux IoT system .
Of the Internet which makes communication with the physical world or end users.
The major application of edge computing is it reduces the transmission cost of bulk volume of data to be transported in resulting traffic and for required distance data must travel, it helps in shrinking latency and improving quality of service. It shortens the distance between the user and the server due to Computation offloading for real-time applications. It eliminates core computing environment, removing a major bottleneck and a potential single point of failure.
VII. GATEWAY THAT CONNECTS CLOUD TO WSNs
As technology grows it will result in the growth of IoT systems for simplicity and billions of devices are connected through IoT gateway
[image: Image result for raspberry pi]
Figure 6- Raspberry pi .
Before sending the data forward the gateway translates between sensor protocols and accumulates sensor data. There are many protocols for connecting a sensor mote to transfer the data to the cloud such as, Wi-Fi, Bluetooth, Ethernet, ZigBee etc. .
Figure 7- Block diagram showing cloud operation.
VIII. CLOUD CONNECTION USED IN WSN
In the past when there was not much development in the radio links, the sensors were connected using wired Ethernet connection internet server as many places did not had the wired connection. Later the cellular networks were introduced and were widely available radio link to maintain a connection between the sensors and the cloud. The sensors needed the wired connection which was an expensive task. After the availability of free radio bands of frequency 902-928 MHz and 2400-2483 MHz frequencies which are accepted as the IEEE 802.15.4 standard which was established in the year 2003. In the year 1997 a new frequency band of range 2400-3483 MHz and 5130-5835 MHz was introduced which was a licence free band which was created according to the 802.11 Wi-Fi standard.
Figure 8- Block diagram showing router operation .
These frequencies were basically used for Wi-Fi access points that were globally used these days. The router was a better means of connection to the cloud and is connected using an Ethernet cable. The mobile phone applications can be connected to the cloud by connecting them to the Wi-Fi router. In the recent years most of the sensors are directly connected to the Wi-Fi using the gsm module. Currently there are many mobile phone applications which make use of the sensors and require data from them and do not require a direct connection to the Wi-Fi router. This is because there is a direct interaction between the sensor and the module present in the mobile phones. The main parameter is the distance between the Bluetooth device and the mobile phone which must be monitored for effective operation to be carried out. To serve this parameter a new standard for Bluetooth communication was created in 1998 which is 802.15.4 which works for the frequency of 2400-2483MHz and recently a new technology known as the low power Bluetooth technology was created as an advancement in the Bluetooth industry which had major application in the wearable devices .
IX. COMMUNICATION PROTOCOL
The communication protocol that is being used for wireless communication is Bluetooth. The Bluetooth protocol stack has similar architecture as other data communication protocols, for example TCP/IP (Transmission control protocol). The fundamental layers of the Bluetooth protocol stack are presented in the Fig. 9 below.
Figure 9- Bluetooth protocol stack .
The Radio layer defines the frequency bands, channel arrangements and transceiver characteristics for a Bluetooth system. The Bluetooth Radio operates at the 2.45 GHz frequency ISM band. It has a total bandwidth of 83.5 MHz divided in 79 channels where every channel has a bandwidth of 1 MHz.
The Baseband is responsible for establishing Radio Frequency (RF) connections with other Bluetooth devices but also to distinguish between Synchronous Connection-Oriented (SCO) and Asynchronous Connectionless (ACL) packets, which can be transmitted over the same link simultaneously.
The Link Manager is responsible for link security, link set-up and configuration. This includes security aspects such as authentication and encrypting by generating, exchanging and checking of link and encryption keys and the control and negotiation of baseband packet sizes.
The Host Command Interface, HCI provides a common interface between the host stack and the lower, hardware-orientated layers.
The L2CAP also known as the (Logical Link and Adaptation Protocol) adapts upper layer protocols over the baseband and is considered as the data-link multiplexing layer responsible for the transportation of data packets. L2CAP distinguishes between higher-level protocols and segments and reassembles packets
The RFCOMM is a serial cable emulation protocol based on ETSI TS 07.10. It has the same functionality as a 9- pin RS 232 serial port and has a built in scheme for null modem emulation.
The SDP protocol discovers provided services and allows other devices to discover which services are provided by the current device .
X. OPERATING SYSTEM IN WSN
Sensor networks have severe limitations in terms of processing power, memory size and energy, while operating in a communication friendly environment that connects both with the physical world and with other sensor network nodes. The operating system must efficiently manage these resources while providing a programming interface, i.e. allow system developers to create resource-efficient software. An operating system multiplexes hardware resources and provides an abstraction of the underlying hardware to make application programs simpler and more portable. Unlike general-purpose computers, which have settled for a number of semi-standardized hardware architectures, sensor network hardware is extremely diverse in terms of processor architectures, communication hardware and sensor devices. There are several operating systems like TinyOS. Contiki , MANTIS, Nano-RK”,Lite-OS , etc. that have been developed, with each offering a different kind of approach for the fundamental problems. But according to a recent survey Linux operating system is the most preferred and widely used OS for WSN based IOT applications.
XI. DATABASE MANAGEMENT IN WSN
Data storage in a wireless sensor network can be divided into centralized storage, local storage, and distributed storage.
Figure 10- Data storage in WSN 
1) Centralized storage: Centralized storage also called as external storage is one of the most simple data storage strategies. Each node transmits the collected data to base station or the sink node for storage, and we can access the data directly from the station. As a result of unlimited storage space of the base station, data can be preserved for a long time, and data access will not consume the energy of the nodes in the network. The sensor network serves only as a means of data collection rather than data processing, because the user can only get the data in the database from the base station.
2) Local storage: Local storage refers to data being stored in the storage of the node itself, and to obtain data we need to input a data access request that is routed to all nodes so as to obtain relevant data. This strategy will broadcast the query request to the entire network, and each node feedbacks the results according to the query condition. Its advantages are that data storage is simple and the storage process is without any communication overhead.
3) Distributed storage: Distributed storage is a data-centric storage strategy. Under this strategy, data is stored in accordance with the specific storage mechanism, and query request obtains the data according to the specific access mechanism. These mechanisms include: hash map, indexing, query and data request being routed according to certain rules. Now that we have observed how the data is stored, we need something which can process and manipulate the data. Thus the most popular and efficient way to do is using SQL.SQL is a standard language to for storage, manipulation, and retrieving data from a database .
To formulate this paper, we have studied the process of designing the sensor mote extensively referring to the nature and properties of the components with the help of available content. This experience has facilitated us to enhance our understanding of the design of sensor mote.
 Dr. Shu Yinbiao and Dr. Kang Lee, Internet of Things: Wireless Sensor Networks
 Stefano Gregori and Franco Maloberti, High-efficiency power amplifier for wireless sensor networks.
 Fatma Karray and Mohamed W.jmal, A comprehensive survey on wireless sensor node hardware platforms.
 Website provide IOT Solution
https://www.mainflux.com/edge-computing.html [Online, accessed 13 February 2019]
 Web article on IOT
https://internetofthingsagenda.techtarget.com/feature/Using-an-IoT-gateway-to-connect-the-Things-to- the-cloud [ Accessed 13 February 2019]
 Reference article by Guest author Larry Burgess is the wireless technical editor at Voler Systems. https://readwrite.com/2015/10/13/sensor-data-device-to-cloud/ [Online, accessed 13 February 2019]
 https://pdfs.semanticscholar.org/8cef/41f785e26312d754e2159b4ee11aa352188a.pdf/  Prabal Dutta , Adam Dunk 2012. Operating systems and network protocols for wireless sensor networks Computer Science and Engineering Division, University of Michigan, Ann Arbor
 Big Data Management of Wireless Sensor Network – Chih-Chieh Hung, Chu-Cheng Hsieh