EEE4121F: Exam prep

Exam prep

Difference between Time Division Duplexing(TDD) and Frequency Division Duplexing(FDD) in mobile networks

TDD: separates uplink and downlink traffic using the same frequency by assigning them different time slots.

FDD: duplexing technique that uses separate frequencies for uplink and downlink simulatenously.

Difference between TDD and TDMA(Time Division Multiple Access)

TDD: separates uplink and downlink traffic for one connection using timeslots on the same frequency.

TDMA: separating multiple different users by dividing a single frequency channel into separate time slots.

Explain three benefits of using the cellular concept in mobile networks

1. High Capacity: Frequency reuse allows the same channels to be used simultaneously in non-adjacent cells

2. Lower Power: Smaller coverage areas mean mobile devices and base stations emit less power, preserving battery life

3. Scalability: Cells can be dynamically sub-divided (cell splitting) as user density increases.

Explain four approaches to increase network capacity in congested cells

1. Cell Splitting: Subdividing a congested cell into smaller microcells to reuse frequencies more often.

2. Cell Sectoring: Using directional antennas to divide a cell into wedge-shaped sectors, reducing interference.

3. Microcell Zones: Distributing multiple antennas to the edges of a cell (all connected to one base station) to share channels and reduce handoffs.

4. Dynamic Channel Allocation: Borrowing channels from less congested neighboring cells on the fly.

Discuss three major use cases for the 5G network and two requirements for each

1. eMBB (Enhanced Mobile Broadband): Requires extreme data rates (10–20 Gbps) and high capacity/traffic density.

2. URLLC (Ultra-Reliable Low Latency Communications): Requires ultra-low latency (<1 ms) and extremely high reliability (99.999%).

3. mMTC (Massive Machine-Type Communications): Requires massive connection density (up to 1M devices/km²) and extremely low energy consumption.

Discuss three approaches for reaching the high-capacity goal of the 5G network.

1. Small cell densification: Shrinking cell sizes to reuse spectrum heavily across urban locations.

2. More Spectrum Utilization: Tapping into brand new high-frequency bounds (like mmWave)

3. Advanced Wireless Tech: Using massive MIMO, NOMA (Non-Orthogonal Multiple Access), and multi-connectivity.

Specify the ranges and features (capacity/coverage) of the three 5G spectrum bands (Low, Mid, High)

1. Low-band (< 1 GHz): Excellent wide-area coverage and building penetration; lowest capacity/data speed.

2. Mid-band (1 GHz – 6 GHz): The "sweet spot" giving a balanced mix of good coverage and high capacity

3. High-band / mmWave (> 24 GHz): Extremely high capacity and speeds; but very limited coverage (requires line-of-sight) and poor penetration.

Compare deploying 5G in Standalone mode (Option 2) vs Non-Standalone mode (Option 3) regarding deployment cost and network slicing.

1. Deployment Cost: NSA is cheaper and faster because it anchors to the existing 4G EPC core. SA is more expensive as it requires building an entirely new 5G Core (5GC).

2. Network Slicing: NSA cannot fully support native slicing due to 4G core limits. SA uses cloud-native 5GC, allowing full end-to-end network slicing customized for vertical industries.

What are three limitations of 5G that 6G aims to solve by integrating Space, Air, Ground, and Sea networks?

1. Connectivity gaps: 5G struggles to cover oceans and extremely remote/rural areas.

2. Lack of Resilience: 5G relies on vulnerable ground infrastructure; Non-Terrestrial Networks (NTNs) provide backup during natural disasters.

3. Backhaul limitations: Satellite and high-altitude UAVs can serve as flexible, wireless backhaul networks where physical fiber is too expensive to lay.

Discuss four approaches for making network selection decisions in heterogeneous wireless networks.

1. Terminal-centric: The mobile device selects the network automatically based purely on signal strength (RSSI).

2. User-centric: Selection is based manually on user-defined preferences (like prioritizing a cheap Wi-Fi network over cellular data).

3. Network-centric: The telecom operator directs traffic to specific networks to balance overall load.

4. Always Best Connected (ABC) concept: Intelligently combining multiple metrics (cost, QoE, bandwidth) to stay connected to the best possible network

Explain how separating the Mobile Station into a Mobile Terminal (MT) and a Subscriber Identity Module (SIM) ensures personal mobility.

Personal mobility allows users to access their services on any physical device. Because identity, encryption keys, and billing are stored on the removable SIM card instead of hardwired into the phone, a user can simply insert their SIM into a new MT and instantly receive their personal calls and data plan.

What is the difference between Terminal Mobility and Personal Mobility?

1. Terminal Mobility: The ability of physical hardware (the device itself) to maintain connectivity during movement/handovers between access points.

2. Personal Mobility: The ability of the user to maintain a consistent identity and access subscribed services regardless of which device or location they use.

Discuss three approaches for prioritizing handoff calls over new calls.

1. Guard Channel Concept: Reserving a fixed subset of channels exclusively for incoming handoffs.

2. Queueing Handoffs: Putting handoff requests into a buffer to wait for a free channel rather than dropping them immediately.

3. Dynamic Channel Allocation: Temporarily "borrowing" channels from a neighboring, uncongested cell to serve the handoff.

Explain the difference between a forward proxy server and a reverse proxy server.

1. Forward Proxy: Sits directly in front of clients/users (e.g., a corporate firewall), intercepting outbound internet requests to enforce security or cache downloads.

2. Reverse Proxy: Sits directly in front of the origin servers (e.g., inside a CDN), intercepting incoming traffic to balance loads and cache content closer to users.

Explain three functions performed by a reverse proxy server.

1. Reactive Caching: Replicates and serves popular content locally to reduce latency.

2. Intelligent Load Balancing: Distributes incoming web traffic across multiple actual backend origin servers.

3. Security / Transcoding: Acts as a shield protecting the origin from DDoS attacks and handles SSL encryption/decryption.

4. SSL/TLS Termination: The reverse proxy decrypts incoming HTTPS traffic, freeing up valuable CPU cycles on backend origin servers

Contrast the advantages and disadvantages of the three types of server selection mechanisms in CDNs (Application, Routing, Naming).

1. Application (HTTP Redirect): Pro: Fine-grained control. Con: Slow (forces extra TCP connection/round-trip delays)

2. Routing (Anycast): Pro: Fast (no extra setup). Con: Blind to actual server load; routes can change mid-session.

3. Naming (DNS): Pro: Avoids TCP setup delays & highly common. Con: Sufferers from "hidden load" (it routes based on the user's Local DNS server IP, not their real IP).

How does the length of media segments affect performance in DASH video streaming?

1. Long segments (10s+): Improve cache performance and reduce HTTP URL overhead, but cause clumsy, inflexible bitrate switching.

2. Short segments (~4s): Allow for smoother, quicker adaptation to sudden bandwidth changes without buffering, but create high overhead due to a massive number of files.

How does TCP BBR improve latency compared to typical loss-based TCP (like Reno/Cubic)?

Loss-based TCP blindly increases transmission speed until routers drop packets, causing buffers to bloat and creating high latency. TCP BBR actively measures bottleneck bandwidth and Round-Trip Time (RTT) to transmit at the exact rate the pipe can handle, keeping routing buffers entirely empty and latency low.

Explain two negative effects of congestion in computer networks.

1. Increased Latency (Delay): Packets sit queued in router buffers for protracted periods.

2. Packet Loss: Buffers eventually overflow forcing packets to be dropped, resulting in wasted bandwidth and mandatory retransmissions.

What are three factors that have contributed to the lack of internet connectivity for 30% of the global population?

1. High infrastructure costs (laying fiber/building towers in rural terrain).

2. Lack of affordability (high cost of data and smartphones relative to local economies).

3. Lack of relevant localized content and poor digital literacy.

What is the difference between a legacy router's forwarding table and an SDN OpenFlow flow table?

1. Legacy table: Relies exclusively on destination-based forwarding (routing based only on the destination IP address).

2. OpenFlow table: Uses Match-Action rules. It can look at completely different layers (MAC address, IP, TCP/UDP ports) and take customized actions (Drop, Read, Modify, Forward).

Explain the functions of the Data plane, Control plane, and Management plane in Software-Defined Networking (SDN).

1. Data Plane: The physical network hardware (switches) that actually forward and move the packets.

2. Control Plane: The centralized SDN Controller software ("brain") that calculates routing tables and programs the switches.

3. Management Plane: The top-level software where human admins write policy, oversee network health, and define abstract business logic.

Discuss three benefits of Software-Defined Networking (SDN).

1. Global View: Centralized management makes securing and monitoring the entire network drastically easier.

2. Programmable Agility: Traffic routing and policy can be changed on-the-fly dynamically via software.

3. Cost savings: Separating hardware from logic allows companies to buy cheap "white-box" switches instead of expensive proprietary legacy hardware.

How does OpenFlow / SDN account for the controller being a "single point of failure"?

1. SDN solves this via a Distributed Controller Architecture.

2. Multiple controllers are clustered together.

3. If the primary "master" controller goes down, a secondary backup controller seamlessly takes over programming the switches so the network keeps running.

What are the functions of the PISA (Protocol-Independent Switch Architecture) framework?

1. Programmable Parser: Extracts headers from incoming packets.

2. Programmable Match Counters: Matches packet fields against loaded rules.

3. Programmable Action Engine: Applies specific commands (modify, drop, forward) to matched packets.

4. Programmable Deparser: Reassembles the packet for egress transmission.

Explain the difference between External Network Virtualization and Internal Network Virtualization.

1. External Virtualization: Combining multiple physical networks into one, or logically carving one physical network into several independent virtual networks (e.g., VLANs)

2. Internal Virtualization: Simulating network hardware in software on a single physical host (e.g., using virtual switches inside a server to let local Virtual Machines talk to each other).

Compare 5G Standalone (Option 2) vs Non-Standalone (Option 3) advantages

1. Standalone (Option 2) Advantage: Complete architectural independence from legacy networks; fully unlocks the cloud-native 5G Core (5GC) for ultra-low latency (URLLC), edge computing, and end-to-end network slicing.

2. Non-Standalone (Option 3) Advantage: Massive reduction in initial Capital Expenditure (CapEx) and faster time-to-market because operators leverage their existing 4G LTE radio access network and Evolved Packet Core (EPC)

Discuss three limitations of the 5G network that will be enhanced by the proposed integration of terrestrial and non-terrestrial networks in 6G.

1. Geographic Coverage Gaps: 5G is heavily restricted to urban/populated landmasses; 6G satellite integration eliminates dead zones over oceans, deserts, and polar regions.

2. Lack of Infrastructure Resilience: Ground-based 5G towers are vulnerable to natural disasters; 6G airborne and space-borne nodes provide immediate backup connectivity when ground grids fail.

3. High-Speed Mobility Support: 5G struggles with severe Doppler shifts and rapid handovers for fast-moving platforms; 6G non-terrestrial tracking smoothly maintains links for supersonic aircraft and high-speed rail lines.

Explain how resources like bandwidth, power, time, and space affect the efficiency of mobile communication networks

1. Bandwidth: Limits the maximum data rate (Shannon's Capacity). Higher bandwidth allows more channels but increases noise vulnerability.

2. Power: Controls cell coverage boundary and Signal-to-Interference-plus-Noise Ratio (SINR). Higher power improves signal quality but increases co-channel interference for adjacent cells.

3. Time: Dictates slot duration in TDMA systems. Precise time alignment/synchronization is required to prevent overlapping slot interference.

4. Space: Determines physical antenna separation and cell layout. Leveraged in space-division multiplexing (MIMO) or sectoring to boost spatial frequency reuse.

Explain two core disadvantages of failing to use harmonized frequency bands for cellular networks.

1. Loss of Global Roaming: Mobile devices cannot operate seamlessly across different countries or regional operators if hardware must support fractured, non-overlapping bands.

2. Increased Device Cost & Complexity: Manufacturers are forced to design complex, multi-band RF front-ends and internal filters to catch disparate frequencies, killing economies of scale.

Using a diagram concept, describe how network management can be achieved by a network operator:

1. Network Management System(NMS): Centralised platform where the operator monitors and controls the network. Communicates with network elements via SNMP, NETCONF, or RESTCONF.

2. Network elements(routers, switches, base stations): Each has a management agent that collects, performance data (MIBs) and responds to NMS commands.

3. FCAPS functions: Fault Mgmt (detect/isolate faults), Configuration (provision devices), Accounting (usage tracking), Performane (QoS monitoring), Security (access control).

4. OSS/BSS: Operations Support Systems (network ops) and Business Support Systems (billing, customer care) sit above the NMS layer.

What is the difference between a forwarding table (legacy routers) and an OpenFlow flow table (SDN switches)?

1. Legacy destination-based forwarding:

a. Matches only on destination IP address (longest prefix match)

b. Populated by distributed routing protocols(OSPF, BGP) running on the router itself

c. Single action: forward out a specific interface.

2. OpenFlow flow table:

a. Matches on up to 12+ header fields (src/dst IP, MAC, port, VLAN, protocol, etc.) - more general.

b. Populated by a centralised SDN controller via the OpenFlow protocol- no distributed routing needed.

c. Multiple possible actions: forward, drop, modify headers, flood, send to controller.

d. Enables fine-grained traffic engineering and policy enforcement not possible with destination-only forwarding.

How does OpenFlow account for the issue of the controller being a single point of failure?

1. Controller clustering/replication: