Home Security System HSS Computer Science Essay

Published: Last Edited:

This essay has been submitted by a student. This is not an example of the work written by our professional essay writers.

Security is a serious issue in today's enterprise environment. The crime rate is increasing due to the successful attempts made by intruders and hackers. In today's life, people are busy catching up their tight schedule and it is common to leave house unattended [2]. Rapid growth of computer control technology and electronic information, led to an idea of intelligent home. Safe home technology enables the user to manage house remotely and improves the comfort of living. When you are in vacation, these systems will send you alert messages and provide immense help at times of emergency. These systems are capable of providing light control, heat control and security.

In this article we will take an in-depth look at various security mechanisms of smart house. We mainly deal with mechanisms like perimeter-based, GSM/GPRS, image processing & 3G based, robot based on sensor network, neural networks using wireless LAN, Bluetooth and ZigBee technologies.

Bluetooth technology which is used to communicate control is found to have disadvantages like high power consumption and high cost. Latest ZigBee technology is used to overcome these problems as it can put devices at reduced level of functionality [9]. They are sent to sleep and wake up states as per the commands given to them, which in turn save power.

Perimeter-Based approach:

Successful implementation of non-functional requirements like performance and security by a home security system is essential. In order to achieve both of these non-functional requirements, perimeter-based home security system is implemented [6].

The different types of HSS used are Wed-Based System (WBS), Passive System (PS), Phone-Based System (PBS) and Hardware-Based System (HBS). In this perimeter-based approach, the house is divided into different perimeters and different HSS is applied in each perimeter in order to achieve non-functional requirements [6].

The house is divided into three perimeters. They are namely P1 includes highly valuable areas of home like office room, bedrooms; P2 includes the remaining area inside the house and P3 includes the remaining area outside the house [6].

Fig 1: Perimeter-Based Security for the house [6]

HBS, WBS is implemented for P1 and P2 respectively. PBS and PS are implemented for P3 [6].

Hardware-Based System:

HBS ensures high security and gives least control to the owner as it is used to guard valuable areas in the home. As soon as sensors are triggered, this system takes decision promptly by triggering actuators without informing the owner [6].

Fig 2: Hardware Based System [6]

Web-Based System:

The two important components of WBS are web server and home controller. Home controller is wirelessly connected with sound, door and motion sensors [6]. When any of these sensors are triggered, then alarm is generated. The cause of the alarm is captured by the camera and can be viewed by the owner on web browser of any device like laptop, phone and personal digital assistant [6]. This is how WBS is said to provide better control to the owner.

Fig 3: Web-Based System [6]

Passive System:

When someone tries to open the door, the sensor associated with the door is triggered. The home alarm connected to the door sensors will make a call to independent monitoring agency or owner using public switched telephone network [6].

Fig 4: Passive System [6]

Phone-Based System:

This is same as passive system. But independent monitoring agents or owners can confirm if it was a false alarm or true one using phone network. They can use the X10 protocol by telephone controller to turn the outdoor light on [7]. If this switches on the light detector that is present inside, door sensor is triggers the alarm again [6]. By this they can confirm that the initial alarm is not false.

Fig 5: Phone-Based System [6]

All these four types of HSS can be summarized using the below tabular form.

Table 1: Different HSS types [6]

GSM/GPRS based approach:

The need for low power consumption and low cost led to the design of Global System for Mobile Communication/ General Packet Radio Service (GSM/GPRS). This includes two parts. They are namely, wireless security sensor nodes and GSM/GPRS gateway [8].

The main features of GSM/GPRS are low cost, low power consumption, easy installing, rapid response, friendly user interface, information security method and emergency alarm function [8].

Fig 6: Structure of the system [8]

GSM/GPRS gateway and network are connected via GSM/GPRS module [8].

This system uses 8-bit PSoC (Programmable System on Chip) microprocessor. The main purpose of using this chip is, it is easy to design, low cost, low power consumption and reduces the design complexity [8].

Fig 7: GSM/GPRS gateway [8]

Microprocessor I receives security information from security nodes through transceiver module. It displays alarm information on LCD and sends out alarm messages to remote users via GSM/GPRS module [8]. Microprocessor II receives the keyboard information and sends it to microprocessor I through serial peripheral interface [8].

GSM/GPRS consists of three wireless security sensor nodes. They are namely, infrared security nodes, door security nodes and fire alarm nodes [8].

Door Security Node:

This security node does not need external power supply as it adopts magnetic sensor. It has ON/OFF signal and its output depends on the distance between the magnet and dry spring [8].

Fig 8: Door security node [8]

Infrared Security Node:

This sensor is placed in different areas like entrances of room. If an intruder enters, sensor will detect the infrared radiation generated by the intruder as the node adopts pyroelectric infrared sensor. The output generated is a small signal and it should be processed. By using microprocessor which is configured of amplifier, filter and analog to digital converter is used to process the signal, thereby reducing power consumption [8].

Fig 9: PIR sensor [8]

Fire Alarm Node:

This security node comprises of temperature sensor and infrared receiver. Temperature sensor is responsible of monitoring environment temperature and infrared receiver is responsible for sensing heat and flames. When these sensors detect that the environment exceeds the defined threshold values, then this node sends out an alarm signal [8].

Image processing and 3G based approach:

For reducing false alarm rates and to have recorded evidence of crime for later inspection, image processing and 3G communication technology is proposed [10]. It achieves this based on ARM9 technology, S3C2410 hardware platform and embedded Linux OS.

The system includes door, infrared, smoke and gas sensors. In case of emergency CPU receives a warning signal and it determines the type of alarm. It controls the camera to the capture video images and stores them in buffer. Embedded platform will preprocess the image data which is converted into file format and this user datagram protocol (UDP) packet is sent through 3G to the user [11]. This is how the owner uses 3G to handle the situation at home.

Fig 10: System's working principle [10]

Fig 11: System hardware design [10]

ARM master module:

ARM is the main module which controls alarm input & output I/O ports, receives and transmitting video data and 3G network communication. It constitutes 64MB high-speed dynamic access memory [10]. The chip S3C2410 consumes low power and has high performance.

Video processing module:

For video processing Video4Linux is used which provides interface functions for video equipment functions.

Alarm input and output circuits:

The alarms and detectors are connected to the controller through General Purpose Input/ Output (GPIO). The system uses photoelectric isolation chip TLP521 to reduce the high power impact of the equipment on the system [10].

This 3G based system has features such as low cost, high reliability, low loss, provide video evidence, high performance and can reduce false alarm rate.

Home security system using mobile robot and sensor network:

A Sensor network based Home Security system (HSSN) has been proposed which is configured by sensor nodes including radio frequency (RF), ultrasonic, temperature, light and sound sensors [1].

This system consists of Home Security Mobile Robot (HSMR), Home Server (HS) and Sensor Network (SN) that is configured by Cricket nodes (CN) [1].

Architecture of the system:

Fig 12: HSSN system architecture [1]

HS uses RF and wireless LAN to receive and transmit home security information. When it receives alarm event information, it transmits event log message to user interface device (UID) [1]. At this point of time, based on the information of each sensor node mobile robot follows optimal trajectory path and reaches the alarm zone (AZ). Then HSMR captures the current context-aware images of alarm zone and transmits then to the user through HS [1]. This is how the user recognizes his home security. The communication among these units of the system is through RF indoor and wireless LAN outdoor.

Fig 13: HSSN system procedure [1]

In SN there are three system operation modes. They are namely, Event Sensing Mode (ESM), Communication Mode (CM) and Home Security Activation Mode (HSAM) [1].

The two procedures in HSSN system are Security Mode Transition Procedure (SMTP) and HSSN Activation Procedure (HSSNAP) [1].

In SMTP, HS receives context-aware information in hop-by-hop through RF from a CN which detects an alarm event from a sensor module. Then HS sends event log message to remote user and informs him about the home security information. At this point of time, all CNs transfers into CM from ESM as wireless communication have to be initialized [1]. At the end of this procedure, all CNs transfers into HSAM from CM as per the commands given by HS. During this entire procedure all CNs activate RF and ultrasonic sensors.

While in HSSNAP, with the help of instructions from HS mobile robot plans an optimal trajectory and reaches the destination. It then captures context-aware images and sends then to HS. HS finally integrates the information that it receives from all CNs and mobile robot and sends them to the remote user using wireless LAN [1]. This helps the user to respond when his house is in dangerous context.

Functions of Sensor network, Home server and Mobile robot:

Sensor network:

The MTS310CA sensor board is equipped with sound, temperature and light sensors. The CN transmits context-aware information which it acquires through the sensor board [1].

Sensor module having sound, light and temperature sensors are added on the cricket module [1].

Fig 14: Cricket hardware unit and sensor board [1]

Home server:

Home server is the heart of the entire system. Based on the information received from CNs and HSMR, HS is capable of determining various security situations like sound, intrusion, fire, light and many more [1]. HS receives information about intrusion and gas detection through wired sensors. HS communicates with UID, Sensor Network and HSMR by wireless LAN, RF and Bluetooth [1].

Fig 15: Wireless communication among HS, HSMR and Personal Digital Assistant (PDA) [1]

The user interface screen of Personal Digital Assistant (PDA) is shown in Fig 5.

Fig 16: User interface screen on PDA [1]

Mobile robot:

In this system, both triangulation algorithm and dead-reckoning method are used to calculate the absolute and relative positions of the mobile robot. For avoiding obstacles and for obtaining optimal path, sensors are also added to mobile robot.

HSMR undergoes three methods. They are namely, Triangulation, Dead-Reckoning (DR) and Path-Planning.

In triangulation, mobile robot uses RF and ultrasonic signal to find the absolute position and distance from the Cricket nodes [1].

Dead-Reckoning is a process in which relative position of mobile robot is determined using two optical encoders and one electric compass [1]. Several measures are taken to reduce the localization error.

Fig 17: Triangulation and Dead-Reckoning process [1]

For path planning, both HS and HSMR share the same home map table (HMT). Mobile robot moves according to the correlation between HMT and the target position.

Fig 18: Path planning [1]

Neural Network approach:

Most of the security mechanisms provided by smart home lack the intelligence for high decision making. This may lead to causing false alarms. Moreover the environment of each smart home varies from place to place and from country to country [2]. So in such case, predefined security instructions may not work well. There arises a need for smart home that can adopt an intelligent algorithm with self-learning capability [2]. Intelligent algorithm such as neural network is capable of evaluating the environment and taking necessary decisions.

There are some existing problems that the security systems fail to verify. When the occupant existed in the house, he will disarm the sensors in the house. So there is no security when the inhabitant existed in the house [4]. Frequent occurrence of false alarm would reduce the efficiency of emergency response. Most of the security systems take long time to react to the triggered alarm and there is a scope for the intruder to escape [5]. This is because these systems have no intelligence in judging the situation and identifying false alarm. Therefore the need for neural network has been recognized in smart home for data analysis.

As training period is long using Backpropagation Neural Network (BPN), neural network is trained using resilient backpropagation algorithm (RPROP) [3].

The training stage of neural network:

Passive infrared (PIR), closed-circuit televisions (CCTV), radio-frequency identification tags (RFID) and readers, magnetic contacts, glass breakage sensors, the time when sensors triggered is fed as input to motion sensors [2]. As numerical data is the only data type that is accepted by neural network, the above input values are converted into numerical data using converter [2]. Later the neural network is trained via hidden neurons and the output of this would be intruder detected or not detected. The desired result is compared with the actual result and hidden neurons are adjusted to improve the system's accuracy. This process is repeated iteratively on hundreds of dataset to increase the robustness of the system [2].

The figure below shows the various steps involved in training phase of neural network.

Fig 19: Training stage of Neural Networks [2]

Cross-Validation stage of neural network:

This stage helps to evaluate the training performance of the neural network and stop the training if needed [2]. The desired output is compared with the actual output to measure the performance. If the performance of the system is good, training is stopped or else it is trained with the next dataset.

Fig 20: Cross-Validation stage of Neural Network [2]

Testing stage of neural network:

After completion of training and cross-validation phases, the neural network is tested with a new dataset from real world scenario. The output of the system is compared with human being's result and later it is used as real world security system if it is proved to work well [2].

Fig 21: Testing Stage of Neural Network [2]