Communication Protocols in Embedded Systems Questions and Answers

Answer: A

The loser withdraws at the bit where it detects a difference and retries later, so no bandwidth is lost to the contention. Ethernet's older shared-medium scheme destroyed both frames and required a back-off, which is the contrast being drawn.

Answer: B

After five identical bits a complementary bit is inserted, so a long run of the same value cannot starve the receivers of edges. It makes the frame length depend on the data, which is why CAN timing is quoted as a worst case.

Answer: C

Global signalling keeps every node's view of the traffic consistent, and the transmitter then retries automatically. Persistent errors raise a node's error counters until it takes itself off the bus, which is the bus-off state.

Answer: D

The counters rise on errors and fall on successes, so a node that is persistently faulty removes itself and stops disrupting the bus. Recovery requires software intervention, which is why a driver must handle the condition rather than ignore it.

Answer: A

Differential signalling over a terminated twisted pair is what survives that distance and that noise, and both candidates are designed for multi-drop wiring. I2C and SPI are board-level buses and USB is a short-range host-centred bus.

Answer: B

A terminated twisted pair with up to thirty-two or more nodes replaces RS-232's ground-referenced point-to-point link, which is what extends the range to a kilometre. The character framing above it is usually still that of a UART.

Answer: C

Leaving the driver enabled blocks every other node from replying, so the turnaround has to happen as soon as the last bit has left the shift register rather than when the software queued it. Getting that timing wrong is the classic RS-485 bring-up fault.

Answer: D

The gap rather than any byte value marks the boundary, which means the receiving software must time the silence and cannot simply look for a delimiter. Modbus ASCII uses explicit start and end characters instead, which is easier to implement and slower.

Answer: A

Correct answer: a unit identifier in the first byte of the frame

Every Modbus frame begins with the unit identifier, and only the slave holding that address responds. The addressing is therefore entirely logical: physical position on the bus is irrelevant, and each slave’s address is configured in the device. Modbus TCP retains the same field even though the IP address has already selected the machine, which is how it addresses devices behind a gateway.

Answer: B

A device never speaks until polled, which is why two devices cannot be connected together and why a microcontroller acting as a host needs quite different software from one acting as a device. On-the-go negotiation exists precisely to swap the roles.