| 1. |
- Cantoro, R., et al.
(författare)
-
Automatic generation of stimuli for fault diagnosis in IEEE 1687 networks
- 2016
-
Ingår i: 2016 IEEE 22nd International Symposium on On-Line Testing and Robust System Design, IOLTS 2016. - 9781509015061 ; , s. 167-172
-
Konferensbidrag (refereegranskat)abstract
- The IEEE 1687 standard describes reconfigurable structures allowing to flexibly access the instruments existing within devices (e.g., to support test, debug, calibration, etc.), by the use of configurable modules acting as controllable switches. The increasing adoption of this standard requires the availability of algorithms and tools to automate its usage. Since the resulting networks could inevitably be affected by defects which may prevent their correct usage, solutions allowing not only to test against these defects, but also to diagnose them (i.e., to identify the location of possible faults) are of uttermost importance. This paper proposes a method to automatically generate suitable test stimuli: by applying them and observing the output of the network one can not only detect possible faults, but also identify the fault responsible for the misbehavior. Experimental results gathered on a set of benchmark networks with a prototypical tool implementing the proposed techniques show the feasibility and provide a first idea about the length of the required input stimuli.
|
|
| 2. |
- Cantoro, R., et al.
(författare)
-
On the diagnostic analysis of IEEE 1687 networks
- 2016
-
Ingår i: Proceedings - 2016 21st IEEE European Test Symposium, ETS 2016. - 9781467396592
-
Konferensbidrag (refereegranskat)abstract
- The IEEE 1687 standard describes reconfigurable structures allowing to flexibly access the instruments existing within devices (e.g., to support test, diagnosis, calibration, etc.), by using configuration modules which act as controllable switches. The increasing adoption of this standard requires the availability of algorithms and tools to automate its usage. The resulting networks might be affected by defects preventing their correct operation. This necessitates the availability of solutions which allow not only to test against defects, but also to identify the location of possible faults via diagnosis. This paper for the first time addresses the problem of the diagnosis of IEEE 1687 networks. Experimental results gathered on a set of benchmark networks show the feasibility of the solution and provide a first idea about the length of the required input stimuli.
|
|
