Birla Institute of Technology & Science, Pilani- KK Birla Goa Campus. Ml Semester 2012-2013 EEEG627 Networked Embedded Applications Test 1 (Closed Book) Date: 15-02-2012 ‘Max Marks: 25 Duration: 1 Hour Q1. A WSN is deployed in a prison where there are 100 cells in two floors all cells facing east with a corridor. The WSN monitors the environment inside in prison cells and also the presence of prisoners in the cell and along the corridors. The prisoners also wear a mote as an arm band. Prisoners are mobile. The cells are single occupancy they are about 10x10x8(| x w x h) feet. The condition inside cells have to be sent at regular intervals to the central monitoring system inside the prison. Also the position of every prisoner inside a cell or outside a cell has to be monitored. {a) What is the deployment pattern — how many motes will be required, are all motes of the same type (classify requirements as motes (sensing, relay, sensing and relay) and super motes, where will you place each mote/super mote? Justify. (b) Would a time synchronization Algorithm be required? If so which one will you use and why? (c) Would a Localization Algorithm be required? if so which one will you use and why? (d) Would use Clustering? If so which algorithm and why? {e) Which Routing Algorithm will you use and why? [Note: If existing algorithms are non-ideal for this application (Q1b — Q1e) you will have suggest the changes that will be required] [20] Q2. For the following Localization Algorithms what are the main characteristics — in terms of (i) Distance Estimation Method (ii) Coarse/Fine Grained? (iii) Range based or Range Free (iv) Centralized Vs. Distributed (v) Time for convergence ? The Localization Algorithms are (a) WCL (b) SWEEPS (c) APIT. Elucidate. [6] Q3. Consider the network shown belowTriangles represent anchor nodes and circles sensor nodes, The dashed lines show the pattern (the network is wireless).All nodes have to be localised. Develop an algorithm for the and show its execution steps. 5) Q4. — Consider the nodes shown below Oo © The nodes need to be synchronized with each other. Assuming that A has a perfect clock {a) How many messages have to be exchanged in case of TPSN (assume that level discovery is completed)? Show the message exchange. (b) How many messages are exchanged in case of RBS? Show the messages? (c) What is the difference in synchronization of TPSN and RBS? (d) Can RBS be used if MAC protocol is TDMA? Elucidate (e) One more neighbour D is added to the scenario such that all nodes are within communication range. How many messages are exchanged for time sync in RBS and how many in case of TPSN? (f) Now add one more node E in communication with all nodes. How many messages are exchanged in case of RBS and TPSN? (g) If the number of nodes is n. How many messages will be exchanged in case of R3S and TPSN? (h) Comment on the scalability of the TPSN and RBS,y Birla Institute of Technology & Science, Pilani- KK Birla Goa Campus Hf Semester 2012-2013 EEEG627 Networked Embedded Applications Test 2 (Closed Book) Date: 21-03-2012 fax Marks: 25 Q4. This is in continuation of Q1 from T1 A WSN is deployed in a prison where there are 100 cells in two floors all cells facing east with corridor. The WSN monitors the environment inside in prison cells and also the presence of prisoners in the cell and along the corridors. The prisoners also wear a mote as an arm band. Prisoners are mobile. The cells are single occupancy they are about 10x10x8(I x w x h) feet. The condition inside cells have to be sent at regular intervals to the central monitoring system inside the prison. Also the position of every prisoner inside a cell or outside a cell has to be monitored. (a) For the deployment pattern you have employed clustering — what will be the average number of nodes in a cluster -what is the MAC protocol that you will use for intra cluster and inter cluster communication and why-how compatible is it with your routing and clustering protocol? (b) If Time Sync is required at regular interval how will you integrate the time synchronization with your normal data packet? (c) In your opinion is the final network protocol reactive, proactive or hybrid? Elucidate by defining the characteristics of proactive/reactive networks. Algorithm be required? If so which one will you use and why? [Note: If existing MAC algorithms are non-ideal for this application (Q1a) you will have suggest the changes that will be required] 10] Q2. A Chemical Plant manufacturing hazardous chemicals needs to be monitored and controlled. ‘The automation pyramid of the Plant is shown below Figi: Network Hierarchy ‘System Details The Diagnostic and Sensing done in the automation pyramid is only for the machinery and devices. A separate WSN is available that monitors only the atmosphere in the factory. This WSN is integrated to the top of the pyramid. The entire industry functioning can be monitored remotely via the internet © Control Information size is maximum 8-bits and real-time. Data rate reaches a maximum of IMbps. * Connection with the actuators and drivers is generally half-duplex. * Connection with sensors is full-duplex. * Sensor and drivers are both connected via a wired interface. * Maximum distance between two devices is 10m. The factory floor covers a considerable area, * Every driver/sensor/actuator has a built in physical address. ‘The lowest level of the pyramid should not be accessible from outside. * Sensors are required to have temporal consistency.«The lowest level of the pyramid is connected using proper infrastructure to the second the automation pyramid Design or suggest the required protocols for the Lowest Level of the automation pyramid. Justify your Protocol © The second level of the automation pyramid acts a sensor data collection as well as control data dissemination point «The Information size varies depending upon whether the information is received from top of the pyramid or bottom of the pyramid. The maximum size of the data can be as large as IKB. * Allconnections to CNC/PLC are wired. © Maximum distance between devices is 100m. «Every device has an inbuilt physical address. © PLC & CNC machines can have an individual network address but should not be accessible from outside the factory. «Data rate can be as high as 10Mbps Design or suggest the required protocols for the Second Level of the automation pyramid. Justify your protocol. = The overall control is done using a set of high-end computers that are isolated from the factory floor. They are available in an isolated control room. The communication between the Control Room Computers and the Factory level network is wireless. «The high-end computers are connected to each other using wired connections. © They are connected to the internet through a high-end fiber optic link. «Large amount of information exchange takes place. The connection has to be highly-reliable connection. Data rate can be as high as 100Mbps. Design or suggest the required protocols for the Lowest Level of the automation pyramid. Justify your Protocol. How will the three levels in the automation pyramid be integrated? Show the entire integration design with sufficient explanation. [10] Q3a. Consider the network shown below Circles represent sensor nodes, ‘The dashed lines show the connection pattern (the network is wireless). Give minimal assignment of a single time slot for each node using TDMA, such that no two nodes within two hops of each other share the same slot. Q3b. For the communication network graph shown in the network of Q3a identify the set of nodes that could be coordinators if SPAN protocol is used for topology control. The protocol has to pick up the minimum number of nodes. Show the steps taken by the protocol. (+2)y Birla Institute of Technology & Science, Pilani -KK Birla Goa campus ‘IL Semester 2012-2013 EEEG627 Network Embedded Application Comprehensive Examination Part- A (Closed Book) Date: 01-05-2013 Max Marks: 20 Duration: 1 Hr Qi. Distingui: ig out the characteristics that are unique to Distributed Systems? ay Q2. Will all Distributed Systems be pervasive? Blucidate a) Q3. What are the required characteristics of'a WSN testing platform? Will they all be available in a single development Tool? Elucidate. a Q4. List what are the security problems common to WSNs and VANETS? What is the effect of each security issue on the network? Is a solution possible in either of these areas? (qd) Q5. What are Migratability & Future Proof in Networks? (1) Q6. What is greedy forwarding? Does it find a place only in Wireless Networks? Can greedy Forwarding be used in WSNs? Elucidate. a Q7. “802.11 protocols can provide deterministic services — hence can also be used in industrial networks without any modificatious”. True or False? Elucidate. aw Q8. An industrial network has real-time as well as non-real-time traffic. How do industrial network protocols generally handle this varied traffic? Can this be done in case of wired and wireless networks? Explain with appropriate examples. a (%. Distinguish between Time Synchronization done in WSN, Vehicular Networks and Industrial Networks? a) Q10. What is delay-jitter in Real-Time Communication? a) QI1. What are fault and error containment in vehicular automation? How are they achieved in intra- vehicular Networks? @) Q12. What is the relative temporal consistency problem in hybrid wired/witeless networks? How this problem is be solved? Will this problem exist in purely wireless scenarios? @ Q13. The data from an 802.11 segment in an industrial network is directed to a segment implementing RS485, The data rate of RS485 is about 1 Mbps and the data rate of 802.11 is 2 Mbps. What will haggen if packet of size (a) 20 units of data (b) 512 units of data is to be sent from the 802.11 segment to RS 485 segment? Elucidate. Would the cut-through method or store and forward method work? Elucidate. @) Distinguish between Clustering in WSNs and VANETS 2 tie theclaba) schedule there is no need to attach source or eeDate: Birla Institute of Technology & Science, Pilani II Semester 2012-2013 EEEG627 Network Embedded Application Comprehensive Examination Part- B (Open Book) M: Duration: 2 Hrs 01-05-2013 Qi. A Nuclear Reactor (Fukushima-Daiichi nuclear plant) is located in a Tsunami prone area. The monitoring system of such a plant will have two modes of operation (1) Under Normal Conditions (Regular Mode) (2) When a disaster strikes ~ (Disaster Mode) ‘System Details ‘The nuclear reactor is made up of 20 sectors — of which 5 of the sectors are deemed hazardous and inaccessible to humans. Such sectors are equipped with a WSN - this network is made up of mobile and fixed sensor units — for monitoring radiation levels. The principle of design for nuclear power plants is defense in depth. Elements of the defense in depth philosophy are Redundant Safety Systems ~ Systems Necessary to support the safe shutdown of the nuclear reactor are designed with redundant and diverse back-up systems ‘Automatic Reactor Protection System — Monitor critical parameters of reactor system and automatically intimate shutdown of the reactor when any of these parameters are exceeded. Radiation Containment Barriers — Physical barriers are built for containment under extreme weather conditions. ‘You are to build the Safety & Control Network for this plant ~ Note: This does not contain plant security in terms of entry to restricted area, perimeter monitoring ete. You can assume that there is a separate network available for this purpose — that can communicate to a central controlling station ‘The Safety and Control Network is divided into four Levels- (a) Device Level (b) Machine Level (¢) Sector Level (d) Central Computing Unit Data rate reaches a maximum of IMbps. Connection with the devices are full-duplex, wired and require temporal consistency. Maximum distance between two devices is 25m. The factory floor covers a considerable area. There is no individual device address —there is only a functional address Devices cannot initiate communication or communicate with each other — they communicate to a Master Device Controller that is interfaced to the next level of hierarchy only — devices should not ‘be accessible to any other level of network / or from outside the network. Design! Suggest/ Modify the required/existing protocols for this level of the network. Justify your Protocol in terms of how it satisfies Network Characteristics. The second level acts a sensor data collection as well as control data dissemination point - The Information size varies depending upon whether the information is received from top or bottom of the hierarchy. Maximum distance between. devices is 100m. Data rate limit is 1OMbps, Machine Level Control is done by a series of high end systems with considerable processing power. Each machine is equipped with a camera — the output of the camera is collected, compressed, encrypted and sent to next level of hierarchy. This is not streaming video but image data collected using camera — based on command from higher level hierarchy. Every machine has an inbuilt physical address. Machines eam have an individual network address but should not be accessible from outside the factory. Design! Suggest/ Modify the required/existing protocols for this level of the network. Justify your Protocol in terms of how it satisfies Network Characteristics.Sector Control Network Characteristics e Thesector control is done using a set of high-end computers that are isolated from the’ * Thehigh-end computers are connected to each other using wired high-end fiber optic Finks speed switches. Large amount of information exchange takes place. The connection has to be hight connection, Data rate can be as high as 100Mbps. In Sectors termed as hazardous the Sector Control Unit also interfaces with a WSN. Design/ Suggest/ Modify the required/existing protocols for this level of the network. Justify your Protocol in terms of how it satisfies Network Characteristics. ‘® The overall control is done using a set of super computers that are isolated from the plant. «The data servers are also part of this system. Data Servers have a daily log of the activities in the plant. They are available in an isolated control room. © They are connected to the internet through a high-end fiber optic link. «Large amount of information exchange takes place. The connection has to be highly-reliable connection. Data rate can be as high as 100 Mbps. Design/ Suggest! Modify the required/existing protocols for this level of the network. Justify your Protocol in terms of how it satisfies Network Characteristics. How will the four levels in the network be integrated? Show the entire integration design with sufficient explanation. ‘Wireless Sensor Network © This wireless sensor network monitors the radiation parameters and structural condition in different ‘sectors. © Each sector has about 100 sensors placed all along the structure enclosing the sector. A fixed deployment pattem has to be followed with sufficient redundancy incorporated. Redundancy is desired in terms of the network functioning as well as topology control. © Data is monitored and collected at regular intervals of time. The collection interval of data is generally a multiple of sensing interval. There is also a possibility that in case of a certain threshold Jevel being reached in the sensed parameter reached, data has to be sent immediately. «The network will also have mobile units that are deployed the moment any sector says there is @ leakage. These are a set of 10 mobile units that have to disperse to different areas within a sector collect data and convene back to the sector control point. Design’ Suggest! Modify the required/existing protocols for this level of the network. Justify your Protocol in terms of how it satisfies Network Characteristics. What type of motes would you prefer for this WSN? What will be the WSN architecture? What will be the deployment pattern that you will use? Design the complete WSN protocol stack. How will the WSN be integrated to the plant network? Now assume that there has been disaster — A disaster warning is sounded and the entire Plant is shutdown. Is it possible that the plant can be remotely shutdown using your current network structure? If yes How? If no — which protocols will you modify to make this possible? ‘The National Guard has been activated to evacuate the town and close road and bridges. The nuclear plant is completely shut down. To get a real-time standard assessment- unmanned aerial vehicle (UAVs) drones are immediately launched. The UAV drones are based in autarkic stations. These not only shelter the drones, but also serve as a node between the cross-linked drones and the remote monitoring service. This is where all information converges. Solar modules ensure the electric power supply. Drones are protected by a dome which opens up in case of emergencyBR crcon's platform contains a charging cradle to recharge the batteries of the drone, Drone has a GPS for navigation and an ultrasonic sensor for distance measurement. These autarkic stations are placed within the safety parameter of the nuclear plant. Upon flying over the contamination site, the UAVs hover collect data relay data to a remote point via the nearest autarkic stations. Each UAV has three radiation detection pods- Two pods to collect and identify naturally occurring radioisotopes of lead and bismuth produced from the radioactive decay of atmospheric radon. The third pod processes data from four radiation detectors to help guide the drone toward the source of gamma radiation. UAVs are equipped with all the needed sensors and computer power to control the flight and carry out complicated missions and tasks without a human operator. The system is also equipped with many safety features that both prevents grave mistakes inflight, and secure a safe landing if there should be a technical problem. The OBEs in the UAVs are safety critical — as these UAVs must not crash at any time and must be able to land safely. Fig 1: UAV Drone in Air from an Autarkic Station shown below UAV Commands: Take -off: Releases the drones Position and Hover: Causes the drone to particular geographic location and hover in that position. Return and land: This command tells your drones to stop doing what they are doing and return to autarkic station. Abandon: Drones can sometimes become stuck and unable to return to their controller's drone bay. In such cases, abandoning them to be able to launch a new wave could be a wise decision, or so that be scooped up later : Scoop to drone bay: Causes disabled drones to be re-docked in the drone bay so that can be re- launched. Again, this only works if the drone is within 2,500 meters of station. What will be the kind of Intra-vehicular buses that will be used with the OBEs on the drones and why? What will be the network formed when UAVs hover? Design the complete protocol stack such that data is relayed to the nearest Autarkic Station. Is this network a WSN? If yes, what are the primary differences between the two WSNs that you have developed in this problem? Elaborate. If no, what is the difference between this and a WSN and why is it necessary that it should be different?