Hierarchical design of fire alarm wireless sensor element

M.Sh. Levin
Sen. Res. Scientist, PhD
Inst. for Information Transmission Problems, Moscow
A.V. Fimin
Student,
Moscow Inst. of Physics and Technology (State Univ.)

1. Introduction

In recent years the significance of sensor systems is increased (e.g., [1],[3],[4],[5],[11],[14],[18],[19]). In general, sensor system is a sensor network including the following layers (e.g., [1],[5],[11],[19]): (i) sensors and sensor local networks (sensor subsystem layer), (ii) communication network (transport layer), (iii) control subsystem (information integration, evaluation, decision making and control). In the article hierarchical design for sensor node structure (configuration) is described using a numerical example for fire alarm wireless sensor element. Mainly, several basic approaches have been applied for the design of system configurations, for example: (i) the shortest path problem [2], (ii) multicriteria multiple choice problem ([10],[15]), (iii) hierarchical multicriteria morphological design (HMMD) approach ([6],[7],[8]), (iv) method based on fuzzy sets [17], and (v) AI methods (e.g., [12]). Here HMMD is used. The approach is based on two optimization problems: multicriteria ranking (outranking technique as modification of ELECTRE [16]) and morphological synthesis based on morphological clique problem ([6],[7],[8]). The numerical example involves hierarchical structure of sensor, design alternatives (DAs) for system parts/components, Bottom-Up solving process. Assessment of DAs and their compatibility is based on expert judgment.

2. Hierarchical morphological design approach

Typical description of hierarchical morphological multicriteria design (HMMD) approach is the following ([6],[7],[8]). A design system is considered as a composition of system parts/components while taking into account their compatibility IC. It is assumed: (1) the system has a tree-like structure; (2) quality of the system is integration of two parts: quality of system parts/components and quality of components compatibility; (3) criteria for system components evaluation are monotonic ones; (4) quality of components and their compatibility IC are evaluated upon ordinal scales. The following designations are used: (1) design alternatives DAs; (2) priorities of DAs r=1,...,k where 1 corresponds to the best level of quality; (3) ordinal compatibility for pairs of DAs w=0,...,p where 0 corresponds to incompatibility and p corresponds to the best level of quality. The solving scheme consists of the following:

Stage 1. Building of the tree-like system model..

Stage 2. Generation of DAs for leaf nodes of the system model.

Stage 3. Hierarchical selection of DAs and their combination into composite DAs for more higher hierarchical level of the system model.

Stage 4. Analysis and improvement of the designed composite DAs.

Let S be the resultant composite system consisting of m parts/components. The composition problem is:

Find composite system S=S(1) * ...* S(i) * ... * S(m) consisting of local DAs (one representative DA for each part of the designed system S(i), i=1,...,m) with non-zero quality of compatibility IC for each pair of the selected DAs.

Discrete space of system quality is based on a vector: N(S)=(w(S);n(S)) where w(S) is the minimum of element pair compatibility in S, n(S)=(n₁,...,n_r,...n_k) where n_r is the number of DAs in S with r-th level of quality. Thus we search for non-dominated by N(S) composite solutions. The solving scheme consists of two phases: (1) constructing the admissible composite solutions, (2) selection of composite solutions which are Pareto-effective by N solutions. Generally this combinatorial problem is NP-hard. The solving scheme can be based on approaches [6]: (a) enumerative algorithm or (b) dynamic programming.

3. Applied design example

3.1 System structure

The following hierarchical structure of a wireless sensor node is considered (including DAs, priorities of DAs are shown in parentheses) (Fig. 1):

0. Wireless sensor node S = H*W

1. Hardware H = M*U*Z:

1.1 Microelectronic components M = R*P*D*Q:

1.1.1 Radio R: 10 mw 916 MHz Radio R₁(3), 1 mw 916 MHz Radio R₂(2), 10 mw 600 MHz Radio R₃(2), 1 mw 600 MHz Radio R₄(1).

1.1.2 Microprocessor P: MAXQ 2000 P₁(1), AVR with embedded DAC/ADC P₂(2), MSP P₃(3).

1.1.3 DAC/ADC D: Motorola D₁(2), AVR embedded DAC/ADC D₂(1), Analog Devices 1407 D₃(2).

1.1.4 Memory Q: 512 byte RAM Q₁(3), 512 byte EEPROM Q₂(3), 8 KByte Flash Q₃(2), 1MByte Flash Q₄(1).

1.2. Power supply U: Li-Ion Battery U₁(3), ), Ni-Cd Battery е U₂(1).

1.3. Sensor Z: Russian Smoke Sensor Z₁(1), Japanese Smoke Sensor Z₂(1), American Smoke Sensor Z₃(3).

2. Software W = Y*O:

2.1 Sensor software Y: Zigbee/802.15.4 & Application Y₁(3), SNAP & Application Y₂(1), Ad-Hoc software & Application Y₃(2).

2.2. OS O: None O₁(1), TinyOS O₂(3).

3.2 Criteria, estimates, and priorities

Criteria, estimates of DAs upon the criteria, and the resultant priorities are shown in Table 1. The set of criteria and their weights shown in parentheses is the following: cost C₁ (100), radius C₂ (1), power consumption C₃ (80), , speed/frequency C₄ (1), fidelity C₅ (10), capacity(memory) C₆ (0.5), development duration C₇ (1000). The resultant priorities (multicriteria ranking) are pointed out in Table 1 and in the system structure above (Fig. 1, in parentheses). Table 2 contains compatibility estimates for some DAs. Mainly, estimates are illustrative ones. For components of M and S equal compatibility estimates are considered.

3.3 Composite solutions

The obtained composite DAs for subsystems are the following:

(a) W₁=Y₃*O₁, N(W₁)=(3;3,0,0); (b) M₁=R₄*P₂*D₂*Q₄, N(M₁)=(3;3,1,0).

While taking into account equal compatibility estimates, the following composite DAs for H and S are obtained: (i) H₁=M₁*U₂*Z₁, H₂=M₁*U₂*Z₂, and (ii) S₁=H₁*W₁, S₂=H₂*W₁. The resultant set of system solutions can be studied by the following ways: (1) multicriteria analysis, (2) expert judgment, and (3) additional usage of HMMD. Further, a solution improvement phase can be considered including searching for bottlenecks and solution improvement (by components, by component compatibility). Note in the example the initial combinatorial set of possible design alternatives includes 5184 (4*3*3*4*2*3*3*2) alternatives.

Table 1

Estimates upon criteria and priorities

DAs	Cost	Radius	Power consumption	Speed	Fidelity	Capacity	Development duration	Priority
R1	6	100	10	-	-	4	7	3
R2	5	20	1	-	-	4	7	2
R3	3	150	10	-	-	9	5	2
R4	2	20	1	-	-	9	7	1
P1	5	-	2	133	-	9	5	1
P2	10	-	1	200	-	9	7	2
P3	30	-	2	600	-	8	5	3
D1	2	-	2	150	14	8	7	2
D2	1	-	1	200	16	8	5	1
D3	2	-	2	300	14	8	7	2
Q1	3	-	1	300	-	8	5	3
Q2	2	-	3	250	-	8	7	3
Q3	3	-	1	100	-	8	5	2
Q4	3	-	1	100	-	8	7	1
U1	5	-	-	-	-	7	5	3
U2	1	-	-	-	-	7	7	1
Z1	1	-	-	-	14	7	5	1
Z2	1	-	-	-	12	7	7	1
Z3	3	-	-	-	16	8	5	3
Y1	10	-	-	-	-	8	7	3
Y2	5	-	-	-	-	8	5	1
Y3	3	-	-	-	-	8	7	1
O1	0	-	-	-	-	7	5	1
O2	0	-	-	-	-	7	7	3

Table 2

Estimates of compatibility

DAs	P1	P2	P3	D1	D2	D3	Q1	Q2	Q3	Q4	DAs	O1	O2
R1	3	3	3	3	3	3	3	3	2	3	Y1	2	3
R2	3	3	3	3	3	3	3	3	2	3	Y2	2	1
R3	3	3	3	3	3	3	3	3	2	3	Y3	3	1
R4	3	3	3	3	3	3	3	3	2	3
P1	-	-	-	1	0	2	3	2	2	3
P2	-	-	-	0	3	0	3	2	2	3
P3	-	-	-	2	0	1	3	2	2	3
D1	-	-	-	-	-	-	3	3	2	3
D2	-	-	-	-	-	-	3	3	2	3
D3	-	-	-	-	-	-	3	3	2	3

4. Conclusion

The paper describes hierarchical approach to modular design of fare alarm wireless sensor. This material is a preliminary one. Clearly, it is reasonable to consider other design problems (e.g., redesign, design of sensor network including all its parts/layers). In addition, it is prospective to use design models under uncertainty. The draft material for the article was prepared within framework of faculty course "Design of systems: structural approach", Moscow Institute of Physics and Technology, Faculty of Radio Engineering and Cybernetics (creator and lecturer: M.Sh. Levin) [9]. The above-mentioned course was partially supported by NetCracker, Inc. [13].

5. References

1. F. Akyildiz, W. Su, Y. Sankarasubramaniam, E. Cayirci, Wireless sensor networks: a survey. Computer Networks, 38(4), 393-422, 2002.

2. E.I. Artamonov, V.M. Hachumov, Synthesis of structures for special tools of machine graphics. Inst. of Control Problems, Moscow, 1991.

3. N. Bulusu, S. Jha, Wireless Sensor Networks. Artech House Publishers, 2005.

4. D. Culler, D. Estrin, M. Srivastava, Overview of sensor networks. IEEE Computer, 8, 41-49, 2004.

5. H. Karl, A. Willig, Protocols and Architecture for Wireless Sensor Networks. Wiley, 2007.

6. M.Sh. Levin. Combinatorial Engineering of Decomposable Systems, Kluwer, 1998.

7. M.Sh. Levin, Modular system synthesis: example for composite packaged software. IEEE Trans. on SMC, Part C, 35(4), 544-553, 2005.

8. M.Sh. Levin, Composite Systems Decisions, Springer, 2006.

9. M.Sh. Levin, Course "Design of systems: structural approach", DETC2006-99547, 18th Int. Conf. on Design Theory and Methodology DTM, Pennsylvania, USA, Sept. 2006.

1. M.Sh. Levin, A.V. Safonov, Design and redesign of configuration for facility in communication network. Information Technologies, 4, 63-73, 2006 (in Russian).

2. F.L. Lewis, Wireless sensor networks. In: D.J. Cook, S.K. Das, (Eds.), Smart Environments: Technologies, Protocols, and Applications. Wiley, 2004.

3. J. McDermott, R1: a rule-based configurer of computer systems. Artificial Intelligence 19(2), 39-88, 1982.

4. Netcracker, Inc. http://www.netcracker.com

5. H. Noltmeier, H.-C. Wirth, S.O. Krumke, Network design and improvement. ACM Computing Surveys, 32(3es), Artcile No. 2, Sept. 1999.

6. V. Poladian, .J.P. Sousa, D. Garlan, B. Schmerl, M. Shaw, Task-based adaptation for ubiquitous computing. IEEE Trans. on SMC - Part C, 36(3), 328-340, 2006.

7. B. Roy. Multicriteria Methodology for Decision Aiding. Kluwer, 1996.

8. A. Smirnov, L. Sheremetov, N. Chilov, J.R. Cortes, Soft-computing technologies for configuration of cooperative supply chain. Applied Soft Computing 4, 87-107, 2004.

9. K. Sohraby, D. Minoli, T. Znati, Wireless Sensor Networks: Technology, Protocols, and Applications. Wiley, 2007.

10. V. Zadorozhny, P. Chrysanthis, P. Krishnamurthy, Network-Aware Wireless Sensor Management. Springer, 2007.