(SEM VI) THEORY EXAMINATION 2021-22 REAL TIME SYSTEMS

REAL TIME SYSTEMS (KCS063)

Section-wise Detailed Answers – B.Tech Semester VI

SECTION A

(Attempt all questions – descriptive explanations)

Q1(a) Define Real-Time System

A real-time system is a computing system in which the correctness of operations depends not only on logical accuracy but also on the time at which results are produced. In such systems, tasks must be completed within specified deadlines. Missing a deadline may lead to system failure or degraded performance. Examples include air traffic control systems, medical monitoring systems, and automotive control systems.

Q1(b) Difference between Hard and Soft Real-Time Systems

In a hard real-time system, meeting deadlines is absolutely critical, and failure to meet even a single deadline may result in catastrophic consequences. Examples include pacemakers and aircraft control systems. In contrast, soft real-time systems allow occasional deadline misses without causing total system failure. Performance degrades gradually rather than catastrophically, as seen in multimedia streaming or online transaction systems.

Q1(c) Priority Inheritance and Priority Inversion

Priority inversion occurs when a high-priority task is blocked by a lower-priority task holding a required resource, while medium-priority tasks preempt the lower-priority task. Priority inheritance is a protocol used to solve this problem, where the lower-priority task temporarily inherits the priority of the higher-priority task until it releases the shared resource. This prevents unbounded blocking.

Q1(d) Open System vs Closed System

An open system interacts continuously with its environment by exchanging information, energy, or resources. It adapts to external changes and is flexible in nature. A closed system, on the other hand, has little or no interaction with its environment and operates under fixed conditions. Real-time systems are typically open systems because they respond to external events.

Q1(e) Byzantine Failures

Byzantine failures refer to faults where system components behave arbitrarily and unpredictably, including sending incorrect or inconsistent information to other components. These failures are difficult to detect and handle because the faulty component may appear operational. Byzantine fault tolerance is essential in distributed real-time systems such as aerospace and military applications.

Q1(f) Overcoming Blocking of Lower-Priority Tasks by Higher-Priority Tasks

Blocking occurs when a lower-priority task holds a resource required by a higher-priority task. This problem is mitigated using resource access protocols such as Priority Inheritance Protocol or Priority Ceiling Protocol. These protocols ensure bounded blocking by controlling resource access and temporarily elevating task priorities.

Q1(g) Reasons for Message Delay in Real-Time Communication

Message delays in real-time communication occur due to factors such as network congestion, packet collisions, propagation delays, processing overhead, queuing delays, and synchronization issues. In real-time systems, minimizing and predicting these delays is essential to ensure timely data delivery.

Q1(h) Sporadic Jobs in Real-Time Scheduling

Sporadic jobs are tasks that arrive irregularly but have a known minimum inter-arrival time. Unlike periodic tasks, they do not follow a fixed schedule. Real-time schedulers must account for worst-case arrival patterns of sporadic jobs to guarantee system predictability.

Q1(i) Traffic Shaping vs Traffic Policing

Traffic shaping regulates data flow by delaying packets to conform to predefined traffic profiles, ensuring smooth transmission. Traffic policing, on the other hand, enforces traffic limits by dropping or marking packets that exceed allowed rates. Shaping is proactive and smooth, while policing is strict and reactive.

Q1(j) Features of UNIX Real-Time Operating System

UNIX real-time operating systems provide priority-based scheduling, fast interrupt handling, inter-process communication mechanisms, memory locking, and real-time extensions. These features enable deterministic behavior and predictable timing required for real-time applications.

SECTION B

(Attempt any three – detailed explanations)

Q2(a) Real-Time OS vs General Purpose OS and Working of RTOS

A real-time operating system is designed to meet strict timing constraints, whereas a general-purpose operating system focuses on maximizing throughput and user convenience. RTOS uses priority-based preemptive scheduling to ensure deterministic task execution. Tasks are scheduled based on urgency, and context switching is optimized to minimize latency. General-purpose OS may delay tasks unpredictably due to time-sharing policies.

Q2(b) Fixed Priority vs Dynamic Priority Algorithms and Optimality of LST

Fixed priority scheduling assigns priorities statically before execution, while dynamic priority scheduling adjusts priorities during runtime. Least Slack Time First (LST) is a dynamic priority algorithm where priority is determined by slack time, which is the difference between deadline and remaining execution time. LST is optimal because it always schedules the task with the least flexibility, ensuring deadlines are met whenever feasible.

Q2(c) Priority Inheritance vs Priority Ceiling Protocol

Priority Inheritance Protocol allows a lower-priority task to inherit the priority of a higher-priority task when holding a required resource. Priority Ceiling Protocol assigns each resource a priority ceiling, and tasks can access a resource only if their priority exceeds the ceiling of all locked resources. Priority ceiling provides better predictability and avoids deadlock but is more complex to implement.

Q2(d) VTCSMA Protocol with Example

Virtual Time Carrier Sense Multiple Access (VTCSMA) is a real-time communication protocol that uses virtual time to coordinate access to the shared medium. Each node schedules transmissions based on virtual clocks, reducing collisions and ensuring deterministic message delivery. For example, in industrial control networks, VTCSMA ensures timely data exchange between controllers and sensors.

Q2(e) Concurrency Protocol Suitability

Different concurrency protocols are suitable under different conditions. Priority inheritance is effective for simple systems with limited shared resources, while priority ceiling protocols are preferred in complex systems requiring strict predictability. Non-preemptive protocols may be suitable where blocking time is minimal and system simplicity is desired.

SECTION C

(Attempt any one part)

Q3(a) Importance of Predictability in Real-Time Systems

Predictability is a fundamental requirement of real-time systems because system correctness depends on meeting timing constraints. Techniques to enforce predictability include deterministic scheduling algorithms, bounded execution times, priority-based scheduling, resource access protocols, and static analysis. Predictability ensures reliability, safety, and certification compliance in real-time applications.

Q3(b) Characteristics and Timing Constraints of Real-Time Systems

Real-time systems are characterized by determinism, responsiveness, concurrency, and reliability. Timing constraints include deadlines, release times, execution times, and response times. These constraints must be analyzed and enforced to ensure system correctness and safety.