E-Commerce Infrastructure (Lecture 2) PDF

Lecture 2 2.1 Infrastructure for E-Commerce When one examines a complex system, it is a good idea to break it up into a number of parts where each part has a specific function to perform. These parts may be arranged as a number of layers. This is similar...

Lecture 2 2.1 Infrastructure for E-Commerce When one examines a complex system, it is a good idea to break it up into a number of parts where each part has a specific function to perform. These parts may be arranged as a number of layers. This is similar to the way one may divide a building into several layers each having a function. The bottommost layer is the foundation-a very important part; over this layer one has a floor. Walls are built over the foundation and enclose the floor. The walls, in turn, support the ceiling. Each layer has a function and provides support to the upper layers. E-commerce systems may also be thought of as consisting of many layers, each layer providing a service. Each layer has a specific function and can be described separately. This gives us a logical way to discuss the architecture of e-commerce systems. One possible layered architecture is given in Table 2.1. We have used six layers to logically discuss e-commerce systems. Each layer has a function and supports layers above it. The bottommost layer is the physical layer. By a physical layer, we mean the physical infrastructure such as cables, wires, satellites and mobile phone system. Their common property is that they provide the communication infrastructure for e-commerce. In fact, without high speed, reliable electronic communication e- commerce is not possible. We call the next layer logical layer as it defines protocols (i.e., a set of mutually agreed rules) to communicate logically between computers connected by the physical network. Organizations found it attractive to use a local network within an organization called intranet. The Internet allows anyone to connect to it. It is thus vulnerable to misuse by anti-social elements who break into others' computers and steal or destroy valuable files. Special precautions are required to prevent unauthorized access. This is provided by what are known as firewalls; which guard the intranets of organizations. Many cooperating organizations lease communication lines and create a private network-interconnecting their intranets. Such a private network interconnecting cooperating organizations is known as an extranet. Private network formed by leasing communication lines is expensive. Thus, methods to ensure secure communication on the Internet 1 between cooperating organizations have been designed. This is called a Virtual Private Network (VPN). A VPN using TCP/IP protocol with enhanced security can also be called an extranet. The next higher layer is the network services layer. This provides services on the Internet infrastructure. The Internet is similar to a railway system which is an essential infrastructure for transporting passengers. The physical layer in a railway system consists of the railway tracks, engines and carriages. The logical layer the signaling system which specifies rules to be followed by engine drivers, guards and station masters for orderly use of the tracks by trains. If rules are broken, collisions take place. Network services layer in a railway system provides reservations for passengers, facilities to transport goods, etc. We require languages to compose messages which can be interpreted by computers. Hypertext Markup Language (HTML) and Extensible Markup Language (XML) provide this facility. As the Internet is accessible to everyone there is always the danger of messages and documents being maliciously altered by unscrupulous persons. Thus, there is a need to send 2 messages which are coded using a secret code. It is also necessary to have an equivalent of signing in the electronic medium also. Those requirements are met by the messaging layer. We call the next layer middleman services. They are essentially services provided to e-commerce participants to make their dealings easier. In our railway analogy, a travel agent who books your railway tickets saving you a trip to the railway station provides middleman services. In e-commerce some important middleman services are secure payments using credit cards, imitating cash payments for small purchases. To authenticate digital signatures, we need an authority to certify public keys of individuals and businesses. Value-added networks provide secure electronic transactions among participants. Hosting services provide among other facilities web presence for organizations and electronic catalogues and directories to participants. All the services provided by the layers described above are essential to support our applications, namely, B2C, B2B and C2C e-commerce. This is thus the top layer (namely, application layer) in our layered architecture. 2.2 LOCAL AREA NETWORK A network connecting computers in a small geographical area such as a building, a university campus or a set of contiguous government offices is called a Local Area Network which is abbreviated as LAN. Usually computers connected to a LAN lie within a radius of about 10 km. In order to connect a computer to a network, an additional electronic circuit known as a network interface card (NIC) is connected to it. This card is required to send messages from the computer to other computers in the LAN and also to receive messages from other computers. A computer wanting to send a message will put it in a small memory (called a buffer) in the NIC and continue with its other tasks. The message will have a header giving the address off the destination computer. It is now NIC's responsibility to transmit the message to the specified destination. One of the earliest methods of interconnecting computers is by connecting their respective NICs to a coaxial cable known as Ethernet connection. In order to communicate among computers connected to a common cable over which only one message can travel at a time, there is a need for a mutually agreed rule for use of the communication medium. Such a rule is called a communication protocol. A protocol known as CSMA/CD (Carrier Sense Multiple Access with Collision Detection) is used. When a computer wants to send a message it delegates the task to the NIC by placing the message in NIC's buffer. 3 The NIC listens to the cable to find out whether any signal is being transmitted by it. This is called Carrier Sense (CS). If no signal is detected, it transmits a data packet. Messages take time to travel in the cable. A message may have been put in the cable by a distant NIC which may not have reached the interrogating NIC. The NIC may thus have wrongly assumed that there is no signal in the cable. If by chance another message was already being carried by the cable when an NIC put its message in it, the two messages will collide and both will be spoiled. Thus, the sending NIC's receiver must listen to signals in the cable for a period of time slightly larger than the time needed for a message from the most distant NIC's transmitter to reach it to detect collision. If a collision is detected by a NIC, it sends a jamming signal in the cable so that all NICs connected to it know that a collision has occurred. When this occurs both the NICs which had put messages in the cable must retransmit. This method of accessing the bus and transmitting packets is known as Carrier Sense Multiple Access with Collision Detection (CSMA/CD). It is called multiple access as any of the NICs connected to the cable can send/receive messages being sent in it -of course one at a time. Each message packet has a source and a destination address along with the message. All NIC's receivers monitor traffic in the cable. If a NIC finds that the message is addressed to it, then it acquires the message and stores it in its buffer storage; else it ignores the message. It is possible to broadcast a message to all NICs connected to the cable by using a special code instead of a single destination address. It is also possible in this method to multicast a message, that is, send the message to a subset of NICs. The number of bytes in a data packet is between 64 and 1518. Adding new computers to the LAN is easy as it can be connected to the spare terminals in a hub. This connection is called 10 Base T. In this notation 10 indicates the speed of transmission of messages on the 10 in megabits per second (Mbps). Base indicates that it is base band transmission, that is, pulses corresponding to bits of a message are transmitted as they are, that is, they are not modulated. T indicates that the wire connecting the computers is a twisted pair of copper wires. 10 Base T has now been upgraded to 100 Base T, that is, 100 Mbps twisted pair based LAN using a hub connection. Currently, the speed has been-further increased to 1000 Mbps. Either twisted pair of wires or fibre optic cables are used to interconnect computers. At this high speed, the distance 4 between the hub and the NIC with so called category 5 unshielded twisted pair of wires is 100 metres whereas with fibre optic cable it is over 250 metres. We have seen so far what is known as a LAN segment, that is, a set of computers connected by using one hub typically in one-department. The number of computers in such a LAN segment will be less than 32 and the maximum distance will be restricted to around 100 metres. These LAN segments should be connected to enable communication between departments. For example, a purchase office may have its own LAN with 16 computers and an accounts office another LAN with 24 computers. 2.3 Interconnecting LAN Segments One way of interconnecting two or more LAN segments is by using another hub (called a backbone hub) as shown in Figure 2.4. When computer C11 wants to send a message to C22, the message is received by Hub 1. Hub 1 will transmit the message to all lines connected to it and also to the line connected to the backbone hub, namely, Hub BB. Hub BB in turn will broadcast it to Hub 2 and Hub 3. When Hub 2 receives the broadcast, it puts it on all connections going out of it. One of them is to C22 which will pick up the message as it is addressed to it. All the hubs in this configuration are simple as they merely amplify and retransmit a message received by them on all other connections. In this design all LAN segments are virtually merged as one LAN. Collision probability increases as the number of computers connected to Hub BB increases. 5  As long as the total number of computers connected to the backbone is smaller than about 40, this method is satisfactory. When the total number of computers to be connected using a backbone connection is larger than 40, use of a backbone hub may lead to too many collisions and consequent reduction of speed of communication between computers. In such a case instead of using a backbone hub to connect LAN segments we use what is known as a backbone bridge (Figure 2.5). A bridge has features to examine each message packet which arrives and determines its destination address. It stores a table which tells the addresses of computers on each of the hubs. Using this information the bridge routes the message to the appropriate hub. A bridge thus isolates the LAN segments controlled by different hubs, thereby, eliminating collisions between messages sent in each of the LAN segments. In theory a bridge can connect any number of hubs provided it has enough memory to store address table and message being forwarded. Another advantage of a bridge is that it can control hubs working at different speeds. For example, Hub 1 may be a 10 Base T and Hub 2 100 Base T. 2.4 PUBLIC SWITCHED TELEPHONE NETWORK Public Switched Telephone Network (PSTN) is the telephone networks maintained by governments or companies in many countries mainly to allow telephone communication among their customers. PSTNs were designed to carry telephone conversations. 6 Telephone conversations are converted to continuously varying electrical signals by the mouthpiece of a telephone. Human conversations lie in a frequency range of 30 Hz to 3300 Hz and thus telephone lines are designed to efficiently transmit signals in this frequency range. The data to be sent from a PC to the ISP is however digital, that is, is and 0s normally represented by two voltage levels, +3 volts and 0 volts. It is not possible to efficiently transmit such digital signals on a PSTN. To enable transmission of digital signals on a PSTN, an electronic circuit called a modem is employed. A modem is connected to the PC of a customer and another to the ISP (See Figure-2.8). The modulator converts 0s and 1s to two analog signals of different frequencies, for example, a 0 may be represented by a tone at 800 Hz and a 1 by a tone at 1200 Hz. The modem at the ISP will convert these tones back to 0 and 1. There are two types of modems-one is called built-in modem. This is usually a printed circuit board connected to a PC's motherboard. The other is an external modem which is a box connected to the communication port of a PC. Early modems used to carry digital information at the rate of 1200 bits per second (bps). With advances in digital signal processing modern modems are capable of speeds up to 56 Kbps using the same old telephone lines. 7 2.5 Broadband Connection to Home PC The speed obtainable with dial-up modem connection is too low to download audio and video files using the Internet. Thus nowadays a so-called broadband connection to homes is provided by PSTNs. This technology known as ADSL (Asymmetric Digital Subscriber Link) uses a modem called ADSL modem to connect telephone line at user premises with the local telephone exchange. This technology uses the fact that copper wires used as telephone lines can inherently support a large bandwidth, much larger than that needed for voice communication. In ADSL technology a filter called a splitter is used in both subscriber's premises and at the exchange. This splitter divides the bandwidth into two disjoint parts. The lower frequencies of 30 Hz to 3300 Hz is used for telephone conversation and the higher band 1 MHz to 8 MHz is used or data communication. The arrangement is shown in Figure 2.9. At the customers premises the ADSL modem converts Os and 1s from PC to analog signal in 1 MHz to 8 MHz band. The splitter isolates this from telephone conversations at the lower frequency band of 30 Hz to 3300 Hz. In other words, in the telephone line voice signals and digital communications from PC can be simultaneously sent as they are in two different bands which are widely separated and thus will not interfere with one another. Normally most users download lot more data than they 8 send to ISP. Thus, data speed from home PC to ISP is usually limited to 256 Kbps and from ISP to home PC is around 1 Mbps. 2.6 WIRELESS NETWORKS Wireless networks are becoming important in e-commerce as customers often want to order goods while they are traveling, that is, while mobile. These are called mobile commerce applications or M-commerce and depend on wireless networks. So, e-commerce using mobile phones is becoming very popular. Wireless technology is primarily used to communicate between mobile laptop computers and ISPs connected to a backbone high-speed network (usually fibre optic cable). In order to use wireless communication, a mobile laptop computer should have a built-in wireless transceiver (a combination of a transmitter and a receiver) and the backbone must have a wireless access point with a transceiver to transmit and receive data from the mobile computer (See Figure 2.11). 2.7 MICROWAVE AND SATELLITE NETWORK For high bandwidth communication over long distances microwave and satellite communication systems are used. Typically, microwave links are used to communicate between an organization and an Internet Service Provider for wide band communication. Microwave links use the frequency band from 2 to 40 GHz. The transmitter and receiver should be in a line of sight without any obstacles in between. That is the reason the microwave transmitters and receivers, which are small dish-like structures (See Figure 2.12), are mounted on top of tall buildings in cities. 9  Inter-city microwave links are set up over hilltops. Microwaves are attenuated (i.e., loose their strength) during transmission. For transmission within a city where the distance between a transmitter and a receiver is less than 50 km this is not a problem. If the distance is more we need what are known as microwave repeater stations. A repeater is placed at a distance of around 50 km from a transmitter. This repeater receives the microwave signal amplifies it and retransmits it. The major advantage of microwave transmission is the large available bandwidth for signals permitting data transmission rates up to 250 Mbps. A microwave link can support 250,000 digital channels each capable of transmitting/receiving 1 Kbps. 2.8 Satellite Communication [ Communication satellites are very useful to send and receive high bandwidth data communications between widely separated points. Satellites are parked in a geostationary orbit at 36,000 km above the equator. The speed of a satellite in this orbit equals the speed of rotation of the earth and thus it is stationary relative to earth. Communication satellites are now launched either by launch vehicles (rockets) or by space shuttles. A communication satellite is essentially a microwave relay station in the sky. Microwave signal at a frequency of 6 GHz carrying the data is transmitted to it from an earth 10 station which has a transmitter. The signal travels a distance of 36,000 km to the satellite and is received as a feeble signal by a system called a transponder mounted on the satellite. It is amplified and retransmitted to the earth using a frequency of 4 GHz by the transponder. The retransmission frequency is different as otherwise the strong return signal will mask the signal received from the earth. A satellite has several transponders thus providing enormous data transmission capability at a cost which is competitive to microwave links. The bandwidth that can be handled by a transponder is above 36 MHz which would support 400 digital channels each of speed 64 Kbps. 2.8.1 The unique features of satellite communication links are :  As the distance of a satellite is 36,000 km even electromagnetic waves take time to travel. Thus, there is a delay of about 240 ms between the time a signal is transmitted to the time it is received by a receiver. This delay has to be accounted for in designing systems based on satellites.  Users can install a receiving dish antenna to receive signals broadcast from a satellite at their premises. Satellite digital radio broadcasts and television broadcasts are now available and they can be received by very small dishes (around 20 cm diameter for radio and 1 m diameter for TV reception) as their bandwidth is small. 11   Satellite radio stations provide data links of 64 Kbps for broadcasting digital data from an organization to multiple recipients.  A transmitting station can receive back the signal sent by it and verify whether it has been correctly transmitted and received. If an error is detected, the signal is retransmitted.  Recently transponders in a satellite allow anyone to connect to it. It has thus been improved allowing lower cost transmitting and receiving systems and antennas to be located on the rooftops of cooperating organizations allowing them to operate private networks. These are called VSATs (Very Small Aperture Terminals). The term aperture refers to the diameter of the dish antenna which is around a meter. 2.9 PRIVATE COMMUNICATION NETWORKS Many organizations which require secure communication prefer to set up their own communication systems. For example, many banks have set up their own communication system connecting their branches. Railways, also have their own private network for their reservation system. Private networks are based on leased lines or VSAT network. VSAT networks have the advantage that they can reach remote areas including hilly areas where PSTNs, do not have large bandwidth access networks. Private networks are expensive to set up. Thus, only large organizations can afford them. Sometimes several organizations cooperate and share the cost particularly if they all will benefit from it. 12