Past exam question: 2076 Chaitra, Q2
Open in the Padhaiko appQuestion
Explain the properties of causal, non-causal and anticausal systems with example.
Solution
Definition and Properties of Causal Systems:
- Definition: A system is causal (also termed physically realizable or non-anticipative) if its output at any present time , , depends strictly on present input and/or past inputs for . It has zero dependence on future input values ().
- Impulse Response Property: For a continuous-time LTI system to be causal, its impulse response must vanish identically for all negative time:
- Physical Realizability: All real-time physical communication systems operating on live signals must be causal because hardware cannot anticipate events before they occur.
- Example: An RC low-pass filter: . Since for , it is strictly causal.
Definition and Properties of Non-Causal Systems:
- Definition: A system is non-causal (anticipative) if its output at any present time , , depends on at least one future value of the input signal, i.e., for .
- Impulse Response Condition: An LTI system is non-causal if for some .
- Realization: Non-causal systems cannot be implemented in real time for live streaming data, but are routinely implemented in offline (non-real-time) signal processing, where stored data files (e.g., recorded images, stored audio buffers, recorded seismic data) allow algorithms to inspect future samples.
- Example: An ideal brick-wall low-pass filter with sinc impulse response: , which extends from to , making it non-causal. Another example: a two-point central moving average filter .
Definition and Properties of Anti-Causal Systems:
- Definition: A system is strictly anti-causal (purely anticipative) if its output at any time depends solely on present and future input values (), with zero dependence on any past input values ().
- Impulse Response Property: An LTI system is anti-causal if its impulse response vanishes identically for all positive time:
- Realization: Anti-causal systems are the time-reversed counterparts of causal systems (). They arise in backward time filtering, bidirectional smoothing filters, and image spatial processing where "future" corresponds to adjacent spatial pixels.
- Example: The backward exponential integrator: with . Here for all . Another example: .
Summary Comparison Table:
| Classification | Dependence of Output | Impulse Response | Real-Time Realizability |
|---|---|---|---|
| Causal | Present & past inputs () | for all | Yes (Live physical systems) |
| Non-Causal | At least one future input () | for some | No (Offline processing only) |
| Anti-Causal | Strictly present & future inputs () | for all | No (Time-reversed processing) |
Answer:
- Causal System: Output depends strictly on present and past inputs (). For LTI systems, for (e.g., ). Physically realizable in real time.
- Non-Causal System: Output depends on future inputs (). For LTI systems, for some (e.g., ). Realizable only in offline stored-data processing.
- Anti-Causal System: Output depends exclusively on present and future inputs (). For LTI systems, for all (e.g., ).