JAWAHARLAL NEHRU TECHNOLOGICAL UNIVERSITY HYDERABAD
JAWAHARLAL NEHRU TECHNOLOGICAL UNIVERSITY HYDERABAD B.Tech IV COMPUTER NETWORKS Previous Question Paper (December - 2020 - R16)
Answer any Five Questions. All Questions Carry Equal Marks.
Main
Explain briefly about circuit switched networks and datagram networks.
Answer
A circuit-switched network establishes a dedicated physical communication path between two nodes for the entire duration of a session. Before data transmission begins, a connection must be set up through the network switches, ensuring reserved bandwidth and constant bit rate. The process involves three phases: circuit establishment, data transfer, and circuit disconnect. Because resources are reserved, there is no contention or congestion during transmission, resulting in low and constant latency. However, it is inefficient for bursty traffic as the reserved capacity remains idle if not utilized, and it lacks flexibility if a link fails.
A datagram network, a type of packet-switched network, operates on a connectionless model where each packet is treated independently. Each packet contains the full destination address, and switches (routers) make independent routing decisions for every packet based on the current network state. Because there is no pre-established path, packets may arrive out of order or take different routes to the destination. This approach is highly efficient for bursty traffic and provides robust fault tolerance, as the network can dynamically reroute around failed links. However, it does not guarantee bandwidth or latency, and packets may be dropped if network buffers become congested.
| Feature | Circuit-Switched | Datagram Network |
|---|---|---|
| Connection | Connection-oriented | Connectionless |
| Path | Fixed path reserved | Dynamic, per-packet path |
| Resources | Reserved in advance | Allocated on demand |
| Latency | Constant | Variable |
| Reliability | High (reserved) | Lower (potential congestion) |
Compare and contrast the OSI and TCP/IP reference models.
Answer
The OSI model is a theoretical, seven-layer reference model developed by ISO, while the TCP/IP model is a practical, four-layer protocol suite developed by the DoD. Key differences include:
- Layering: OSI separates the Application, Presentation, and Session layers, whereas TCP/IP combines these into a single Application layer.
- Connection Modes: OSI supports both connection-oriented and connectionless communication in the Network layer, but only connection-oriented in the Transport layer. TCP/IP supports connectionless in the Internet layer and both in the Transport layer.
- Approach: OSI was designed before the protocols were defined, making it a generic model. TCP/IP was designed around existing protocols, making it more practical for real-world implementation.
- Complexity: OSI is often considered more complex due to its strict layering, while TCP/IP is more flexible and widely adopted as the standard for the Internet.
Identify the advantages of optical fiber over twisted-pair and coaxial cable.
Answer
Sign in to view the complete answer paper.
In analyzing ALOHA, we use only one parameter, time; in analyzing CSMA, we use two parameters, space and time. Can you explain the reason?
Answer
Sign in to view the complete answer paper.
Discuss the ATM layers and their functions.
Answer
Sign in to view the complete answer paper.
Briefly discuss subnetting and supernetting. How do the subnet mask and supernet mask differ from a default mask in classful addressing?
Answer
Sign in to view the complete answer paper.
Elaborate why do OSPF messages propagate faster than RIP messages?
Answer
Sign in to view the complete answer paper.
In regard to quality of service, describe how do user-related attributes differ from network related attributes in ATM?
Answer
Sign in to view the complete answer paper.
Explain how can a system prevent a guessing attack on a password? How can a bank prevent PIN guessing if someone has found or stolen a bank card and tried to use it?
Answer
Sign in to view the complete answer paper.