Computer Network Questions and Answers

Answer: B

Correct answer: Link state

In a link state protocol every router floods a description of its own links to all the others, so each ends with the same map of the network and computes its own shortest paths from it. Distance vector routers exchange only their own distance tables and never share a common database.

Answer: A

Correct answer: Hot potato routing

Hot potato routing sends each packet out on whichever queue is shortest, getting rid of it as fast as possible without regard to where it is going. Flooding sends copies everywhere, and static routing consults a fixed table.

Answer: B

Correct answer: time-division

A time-division switch works by writing incoming samples into memory and reading them out in a different order, so each sample must be stored and fetched again. That store-and-fetch is the source of the delay; a space-division crossbar makes a direct path and adds none.

Answer: B

Correct answer: Entropy

Entropy measures the average uncertainty of a source: the more evenly spread its symbol probabilities, the more information each symbol carries. Bandwidth measures the channel, not the source.

Answer: C

Correct answer: Converts \(2^{n}\) input lines into an \(n\)-bit output code

An encoder takes \( 2^{n} \) input lines, of which one is active, and produces the \( n \)-bit binary code identifying which. A decimal-to-BCD encoder, for example, has ten inputs and four outputs. Its inverse is the decoder, described by the second option, which takes an \( n \)-bit code and activates one of \( 2^{n} \) outputs - used for address decoding and for driving displays. A plain encoder assumes only one input is active at a time and produces nonsense if two are. A priority encoder removes that limitation by responding to the highest-numbered active input and ignoring the rest, which is exactly what an interrupt controller needs when several devices request service at once. The third option describes a multiplexer and the fourth a flip-flop. Encoders and decoders together allow a keyboard, with one line per key, to send a compact code over few wires and have it expanded again at the other end.

Answer: A

Correct answer: CMOS

CMOS - complementary metal-oxide-semiconductor - consumes very little power, which is why it displaced every earlier family and is used in essentially all modern integrated circuits. The reason lies in its structure. Each gate pairs a p-channel and an n-channel transistor so that in either stable state one of the two is off, and almost no current flows from supply to ground. Power is drawn chiefly when the output changes, charging and discharging the capacitance of the next stage - which is why CMOS power consumption rises with clock frequency. TTL, transistor-transistor logic, is faster than early CMOS but draws current continuously. ECL, emitter-coupled logic, is the fastest of the three because its transistors never saturate, but it consumes the most power of all. RTL was among the earliest families and is obsolete. CMOS also tolerates a wide supply range and has high noise immunity, though it is sensitive to static discharge.

Answer: C

Correct answer: Carrier sense systems

Carrier sense schemes can be run either way: as plain contention, in which every station has equal standing, or with priority, in which higher-priority stations wait a shorter interval before transmitting and so win the medium. TDM and register insertion allocate access by position rather than by priority.

Answer: B

Correct answer: 20,000 bits

The signal takes

\[ \frac{200\times 10^{3}\ \text{m}}{200\ \text{m}/\mu\text{s}} = 1000\ \mu\text{s} \]

to go round, and at 20 Mbps that interval holds

\[ 20\times 10^{6}\times 10^{-3} = 20{,}000\ \text{bits} \]

Answer: C

Correct answer: 11010

Dividing repeatedly by two and reading the remainders upward gives the answer. 26 gives remainder 0, then 13 gives 1, then 6 gives 0, then 3 gives 1, then 1 gives 1 - reading upward, 11010. The check is quicker than the derivation: the place values from the left are 16, 8, 4, 2, 1, so 11010 is \[ 16 + 8 + 0 + 2 + 0 = 26. \] An alternative method subtracts the largest power of two that fits and repeats: 26 − 16 = 10, 10 − 8 = 2, 2 − 2 = 0, so the bits at 16, 8 and 2 are set. The distractors are all near misses of the kind produced by reversing the remainder order or dropping a bit, which is why reading the remainders in the right direction matters. In hexadecimal 26 is 1A, and grouping 11010 into fours from the right - 0001 1010 - gives 1 and A directly.

Answer: C

Correct answer: a small, very fast storage location inside the processor

A register is a small storage location built directly into the processor, typically holding one word - 32 or 64 bits - and accessible in a single cycle. Registers are the fastest storage in the machine and sit at the top of the memory hierarchy, above cache, main memory and disk. They are few in number precisely because they are expensive in silicon area, and a large part of what a compiler does is decide which values are worth keeping in them. Several have fixed roles. The program counter holds the address of the next instruction; the instruction register holds the instruction being decoded; the accumulator holds arithmetic results in simpler architectures; the memory address and memory buffer registers interface with memory; the stack pointer tracks the top of the stack; and the status or flag register holds the zero, carry, sign and overflow bits that conditional branches test. General-purpose registers hold whatever the program needs.