Developing an Innovative Mobile and Wireless Networks Course

Transcription

1 Developing an Innovative Mobile and Wireless Networks Course Hetal Jasani Assistant Professor School of Technology Michigan Technological University Abstract All over the world, mobile and wireless courses become increasingly popular in colleges (including community colleges) and universities. This paper discusses about developing the innovative course of mobile and wireless networks using laboratory activities. It elaborates innovative projects that are suitable for laboratory work in network engineering (including technology) curriculum. It explores both hardware and software components that are already being used for practical exercises in mobile and wireless network course. Most often these laboratory exercises include both simulations involving modeling software and hands-on projects. It is often impractical to perform hands-on experiments with all different newly developed technologies due to lack of resources in university environment. Therefore, it is attractive to use virtualization environment (modeling software) to perform different laboratory activities. In mobile and wireless network engineering courses, instructors mostly focus on lower protocol layers such as physical, MAC (Medium Access Control) and network (routing) layers. In learning the basic concepts of mobile networking protocols via simulations (modeling software such as OPNET [8]), students have ample opportunities to characterize the performance of mobile and wireless networks. This paper discusses the simulation (virtual) projects for modeling and analyzing the different mobile and wireless network technologies. In addition, this paper also explores different options available, both in hardware and in software, that have been recently used as laboratory assignments and semester projects. I. Introduction The field of mobile and wireless networks is dynamically changing due to the advances in the technologies that support it. Many areas demand highly trained personnel to solve the new challenges such as wireless local area network (WLAN), wireless security, WiMAX (Worldwide Interoperability for Microwave Access), Cellular Networks, Satellite Networks, etc. There is a great demand of technicians and engineers who can maintain and operate these wireless systems. While electrical and computer engineering and technology curriculums offer students few wireless networking courses; this may not enough to train network engineers with the proper background on the newer wireless technologies. Most of these courses are focused on theory and

2 2 not on hands-on experiments that are very useful to gain the understanding of fundamental concepts. The goal of mobile and wireless networks course is to familiarize students with several different wireless networking technologies through a series of laboratory experiments using small-scale test beds and employing computer simulations. The equipments for performing laboratory experiments with the newest technologies are expensive and changes constantly [1,2]. Consequently, it is very difficult to upgrade and continuously enhance these laboratories. As a result, employing a computer based laboratory with different software packages specialized for different mobile and wireless network courses might be a good approach to overcome all of these aforementioned drawbacks. However, hands-on experiments should not be completely eliminated [3]. Many universities have used different protocols and equipments for hands-on experiences in wireless and mobile network courses. The protocols and standards include IEEE (a, b, g or simply WiFi) [5], Bluetooth (IEEE [6]) and WiMAX (802.16) [7]. The CNSA (Computer Networking and Systems Administration) program [9] in the School of Technology [10] at Michigan Technological University [11] offers several courses in networking and system administration. In general, students can learn about many networking systems, but had rare opportunity to face the performance trade offs involved in designing a system using simulation software. In newly designed Mobile and Wireless Network course (SAT2600) with enhanced laboratory experiments have demonstrated effectiveness in teaching the concepts of different wireless network technologies [4]. This course has been proposed to provide a practical view of mobile and wireless systems using computer modeling and simulation combined with real equipment configurations. The course assumes that students only have basic knowledge of networking (i.e., students have taken first course of network administration). Therefore, it includes a lecture covering the relevant concepts needed to understand the different IEEE [5], IEEE [6] and IEEE [7] standards, etc. and their applications. OPNET Inc. [8] has provided complementary software for educational and not-for-profit purposes to enhance these laboratories experiments. Some other WLAN equipments were bought form student lab fees. The laboratory has also intended to help the research activities of faculty members. This paper discusses some of lab experiments belong to mobile and wireless networks course. In the rest of the paper, we focus on specific approaches taken at our CNSA program. Section II discusses the general course development approach. Section III illustrates some sample simulation and hands-on experiments. Section IV discusses the results from the mobile and wireless course in CNSA Program. Section V concludes our discussion on developing mobile and wireless network course. II. Course Development The objectives of mobile and wireless network course in CNSA program at Michigan Tech are that students should be able to: Understand the standards used in wireless local area networks

3 3 Evaluate the performance of mechanism in standard using OPNET Modeler Understand the technologies used in Wireless Metropolitan Area Network (WMAN) Understand and implement the wireless security with modern wireless security solutions Learn to configure the equipments from two of the principal wireless LAN vendors, Cisco and Linksys. For the physical networking experiments, students use several of-the-shelf networking devices such as routers, computers, cables, WLAN cards, Linksys wireless router, etc. To configure and do analysis of wireless networks, the students configure a router, some computers and a wireless access point. Eventually, they create larger models using OPNET software. The high level behavior of different applications in IEEE networks will be simulated on discrete event simulations. This allows students to measure several performance characteristics such as end-toend delay, aggregate throughput, congestion, number of retransmissions, etc. in such kind of networks. Course outline for Lab: The topics are scheduled for a 14-week semester with laboratory assignments. The subjects covered are [12, 13]: Wireless LAN Devices and Standards How Wireless Works IEEE Physical Layer Standards IEEE Medium Access Control and Network Layer Standards Planning and Building a Wireless LAN Conducting a Site Survey Wireless LAN Security and Vulnerabilities Implementing Wireless LAN Security Managing a Wireless LAN Network Settings and Wireless LAN Troubleshooting Personal, Metropolitan, and Wide Area Wireless Networks III. Sample Hands-on and Simulation Experiments Some sample experiments are discussed below to demonstrate the major areas of this course. Initially, students perform hands-on experiment using command line interface to configure the Cisco Access Point (AP) (e.g., set up IP address, SSID, etc.). Students also execute more experiments to increase the security of wireless networks by using MAC filter, IP filter and WEP/TKIP. At last, few simulation experiments are proposed to model a wireless network and use OPNET Modeler software to evaluate the performance of mobile and wireless networks. Students are asked to customize the IEEE MAC protocol (i.e., enabling RTS/CTS, PCF, etc.) using OPNET Modeler. Due to the limitation of space, only four sample labs are described in this paper focusing on simulation based labs.

4 4 A. Sample Lab#1 (Simulation): Effect of RTS/CTS on wireless networks Objectives: This experiment aims to observe the simulation of a mobile and wireless network by utilizing a technique known as discrete event simulation, where a computer replicates the behavior of an event that occurs in the wireless network. In addition the student must customize several wireless clients/nodes/ap to interact in OPNET network model [8]. Simulation model shows the impact of the RTS/CTS mechanism as a measure to prevent the hidden node problem in based wireless network. RTS/CTS is an optional mechanism that may helpful in hidden node problem scenario. Figure 1 shows that Node A and Node C are hidden from each other and can not hear each other since they are out of transmission range of each other. Figure 1 Hidden Terminal Problem Figure 2 Network Topology for Lab#1

5 5 As shown in Figure 2, students place two nodes and one receiver representing a campus-wide ad hoc WLAN. Receiver, here, represents an access point for infrastructure based WLAN. Students draw a trajectory for Node A, on which it move according to the predefined amount of time and distance. Students set up simulation such that Node A move 430 meters (Node A go out of range of Node B) along the path represented in the Figure 2 over the course of about five minutes. Task Assigned: OPNET [8] is a discrete event simulation program employed to analyze the behavior of communication networks. Students use it to understand the behavior of several scenarios in wireless networks. To study the properties of wireless networks, the students had to configure and simulate different scenarios provided with the software tutorial. Students run a simulation on a wireless LAN (WLAN) scenario, which suffers from hidden node problem. They observe the effect of the problem using the collected statistics that delay is increased significantly when nodes are hidden from each other in wireless network. They enable the RTS/CTS feature to overcome this issue. Eventually, students verify that the network performance improves as a result of enabling the RTS/CTS frame exchange in based wireless networks. Discussions: The program generates events based on source generators (create new packets, follow steps in IEEE standard protocol, etc.). Each event has certain properties that are passed from one block to the next. At the end, the program collects statistics associated with each event and they are reported in the final output. The most common parameters are the delay, aggregate throughput, retransmission attempts, etc. In Lab#1 scenario, two nodes become hidden from each other for approximately 300 seconds. During this period, use of RTS/CTS frame exchange reduces the number of collisions (and therefore retransmissions) significantly. CTS message sent by the intermediate node informs the sender nodes about the upcoming data transmission attempt of the other node and its reservation of the channel. Due to less number of retransmissions, the wireless LAN delay drops drastically for the period when the nodes are hidden to each other. B. Sample Lab#2 (Simulation): Planning and Analyzing Wireless Networks using OPNET Objectives: In this lab, students become familiar with WLAN model attributes needed to configure BSSs. They use the model to select an appropriate WLAN topology according to the application traffic such as web browsing, , database traffic, etc. Task Assigned: In this lab, students create an independent BSS wireless LAN network that spans multiple floors on a building in OPNET network model. Appropriate attributes are configured and simulation is run. Different network parameter statistics are collected and students observe that the network performance is not satisfactory. Students add two more access points to build an ESS network in wireless network to see if addition of APs could improve the performance of the network. Total of three APs are used in the network to avoid the interference due to the channel overlapping.

6 6 Another simulation is run to find out whether the additional access points improve the network performance. Figure 3 OPNET Network Model for Lab#2 Discussions: Students observe that deployment of additional access points increased the WLAN capacity. Distributing the wireless clients on different floors among the access points reduced the contention for each shared medium. WLAN packet drops are observed and significant reduction in it is found as positive effect. Moreover, application throughput is increased significantly. Students also observed that average WLAN delay is significantly lower. Students can conclude from the lab that increasing the number of access points is an useful alternative when the data rate cannot be increased. Figure 3 shows OPNET network model for this lab.

7 7 C. Sample Lab#3 (Simulation): Effect of PCF on Wireless Networks Objective: Student use OPNET modeler to see the effect of using PCF mechanism to improve the performance of real time applications over WLAN. Task Assigned: Student use OPNET modeler to observe that with the initial configuration of the wireless LAN (under DCF mode only), the performance of the network is not satisfactory for voice clients. Student enable PCF mode on the access point and configure the voice clients as contention-freeperiod pollable stations. Simulation is run to evaluate the performance of the wireless LAN under new configuration for the clients of voice application and other (non-real-time) applications. Students observe the different network characteristics to solidify and conclude the effect of PCF in based wireless networks. Figure 4 OPNET Network Model for Lab#3 Discussions: In PCF enabled scenario, voice application shows a better performance compared to the previous scenario in which only DCF was used. The end-to-end delay for voice packets has dropped considerably in PCF enabled scenario. Students also observe significantly less number of occasional high delays, which reduced the measured average voice jitter value below the half of

8 8 the average jitter value from only DCF operation. This should increase the voice quality for the voice application clients. Average response times for HTTP and FTP application clients are observed and students observe positive results. Students can conclude that for this particular network and application traffic pattern, deployment of PCF improves the performance of Wireless LAN for the client of all application types. D. Sample Lab#4 (Hands-On): Bluetooth interference Project Objective: Students identify the potential interference between Bluetooth and b by performing the hands-on experiment using total of four wireless clients in closed proximity. Task Assigned: In this lab, one team of students set up a Bluetooth Piconet. Another team set up the b network. First team will install the Bluetooth utility and adapter on two clients to communicate with each other using IEEE standard. Another team will setup the ad hoc mode wireless network [14] between two wireless clients in close physical proximity to Bluetooth machines. In first experiment, one b client will download file from another b enabled device. Second time, same experiment will be repeated while performing download among Bluetooth devices (transferring file from one Bluetooth device to another Bluetooth device). Discussions: While performing second download, Bluetooth enables devices will create interference to b devices since they all operate on the same frequency. Students observe that throughput go down while performing second experiment and comparing it with first one. Please note that first part of this lab will be done without Bluetooth devices and hence will be performed without any interference. Students conclude their observation by noting the throughput and hence learn the effect of interference on b network. IV. Results from the Course During the semester, we encountered several resource problems with the operation of the computer laboratory. In particular, the client wireless adapter was not good enough to perform all experiments. It will be upgraded to Cisco wireless adapter which is enterprise level solution for all wireless needs. However, many experiments were performed using Linksys wireless adapter (802.11b based). Students were given extra time, for assignment completion, to balance the resources and accommodate the student needs. This course only has Network Administration-I (SAT first course of networking) as a prerequisite. OPNET was not used in SAT2343. Since the students haven t had any experience with OPNET or any other simulation software, they had some difficulties learning it. It is due to the learning curve involved in OPNET. The lectures were designed to discuss different IEEE standards. At the end of the semester, all of the students were able to complete all the lab assignments.

9 9 V. Conclusions The objective of this paper was to develop mobile and wireless network course using laboratory and project assignments. We presented few simulation and hands-on experiments. The computer simulation laboratory is an economical alternative to deploy a full equipped real environment. However the use of simple components such as those used in small business, permits the students to understand the application of the simulations. Such a laboratory has a significant advantage over a other traditional environments because the capability to upgrade and update software packages and replacement of economical components is much easier and cost effective. We have designed different lab assignments. The lab assignments include RTS/CTS, PCF, planning deployment, etc. Students perform experiments using computer modeling and simulation, submit lab reports, and complete evaluation forms to give a feedback in order to improve and update the assignments for coming semesters. Students find this lecture course along with lab assignments helpful to them in understanding the theory of mobile and wireless networks, gaining practical experience, and learning the performance trade offs involved in setting up optional mechanism such as RTS/CTS in specific networking scenarios. In the future, more lab assignments would be developed to cover advanced topics in wireless network such as network management, enterprise level security solutions, etc. More OPNET labs will be designed to enhance learning the wireless network standard such as: IEEE e, IEEE n. WiMAX lab will be developed to understand the WiMAX technologies and its uses. Only problem is that students of our wireless class are at sophomore levels and they will have hard time to learn OPNET in short period of time such as one semester. VI. References [1] N. K. Swain, M. Swain, and J. A. Anderson, Integration of virtual instruments into an EET curriculum, Firenze, Italy, [2] C. Rosenberg and S. G. M. Koo, Innovative and easy-to-deploy communication networking laboratory experiments for electrical and computer engineering students, Como, Italy, [3] Z. Nedic, J. Machotka, and A. Nafalski, "Remote laboratories versus virtual and real laboratories," Reno, NV, [4] R. P. Ramachandran, L. M. Head, S. A. Mandayam, J. L. Schmalzel, and S. H. Chin, Laboratory experiments unifying concepts in the communications, digital signal processing (DSP) and very large scale integration (VLSI) courses, St. Louis, MO, United States, [5] IEEE, Wireless LAN Media Access Control (MAC) and Physical Layer (PHY) Specification, IEEE Draft Version 4.0, May [6] IEEE WPAN High Rate Alternative PHY Task Group 3a (TG3a), Dec [Online]. Available: last accessed June 29, [7] IEEE , IEEE Standard for Local and Metropolitan Area networks-part 16: Air Interface for Fixed Broadband Wireless Access. (2004).

10 10 [8] Online Documentation, OPNET Modeler, Date visited: March [9] CNSA Program at Michigan Tech, last accessed June 29, [10] School of Technology at Michigan Tech, last accessed June 29, [11] Michigan Technological University (Michigan Tech), last accessed June 29, [12] Kelly Cannon, Lab Manual for CWNA Guide to Wireless LANs, Second Edition, Thomson Course Technology [13] Mark Ciampa, CWNA Guide to Wireless LANs, Second Edition, Course Technology Incorporated, [14] C. K. Toh, Ad Hoc Mobile Wireless Networks: Protocols and Systems, Prentice Hall, December Biographical Information HETAL JASANI Dr. Hetal Jasani is an assistant professor at School of Technology in Michigan Technological University, U.S.A. His research interests include network protocols, distributed systems and network security. He received his Ph.D. in Electrical and Computer Engineering from Florida International University in He is a member of ASEE, IEEE and ACM. Currently, Dr. Jasani teaches Mobile Computing, Network Security, Introduction to Programming, Unix & Linux Administration courses.