WO2013098861A1

WO2013098861A1 - Monitoring system of an intrusion barrier.

Info

Publication number: WO2013098861A1
Application number: PCT/IT2011/000421
Authority: WO
Inventors: Vincenzo DE ASTIS; Alfredo GIGADA; Emanuele ZAPPA; Stefano Manzoni
Original assignee: Cias Elettronica S.R.L.
Priority date: 2011-12-29
Filing date: 2011-12-29
Publication date: 2013-07-04

Abstract

The present invention concerns a security system for monitoring an intrusion detection barrier (2), said security system comprising a plurality of sensors (Mj) arranged along at least part of the perimeter of said intrusion detection barrier and firmly attached thereto, and a microcontroller (3, AP, AC) in signal communication with said plurality of sensors. The system is characterized in that said plurality of sensors (Mj) are MEMS sensors and are divided into a plurality of unique sets (SBSi), each unique set comprising at least one MEMS sensor and defining a zone (2B) of said intrusion detection barrier, each MEMS sensor generating at least one electric signal (Vei) in response to one or more mechanical actions on said intrusion detection barrier (2); said microcontroller (3, AP, AC) being configured to respond to each of said electric signals (Vei) generated by said plurality of MEMS sensors, said microcontroller (3, AP, AC) being configured to detect said mechanical action when said at least one electric signal (Vei) is received, said microcontroller being configured to identify said zone (2B) of said barrier based on the reception of said at least one electric signal (Vei) corresponding to one of said plurality of said unique sets of MEMS sensors, said microcontroller being configured to generate an alarm signal (Vaii) in response to said mechanical action, said alarm signal being indicative of said at least one zone (2B).

Description

Monitoring system of an intrusion barrier.

Technical Field

The present invention relates to a monitoring system for monitoring an intrusion detection barrier, and particularly to a security system for monitoring the perimeter of an intrusion detection barrier to detect any break-in event, as defined in the preamble of claim 1.

Background art

Various monitoring systems are currently available for monitoring intrusion detection barriers, such as metal fences, to detect any break-in event. These security systems generally rely on sensors which sense vibrations propagating over the intrusion detection barrier.

Particularly, the sensors are integrally associated with the intrusion detection barrier and may be implemented in lumped- or distributed-constant arrangements.

In the former case (i.e. with sensors in lumped constant configuration), discrete piezoelectric or electromechanical inertial mass transducers are used as sensors, and are arranged along a cable that is designed to carry information to a common processor.

In the latter case (i.e. with sensors in distributed constant configuration), the sensors are characterized by a single transducer, consisting of the cable itself, which uses the triboelectric effect, or the piezoelectric effect or the capacitive effect, to generate an electric signal proportional to the vibration of the fence on which it is installed.

While the above systems have valuable features, they are still affected by a number of limitations.

For example, the systems that rely on piezoelectric sensors in lumped constant arrangements, generate mV electric signals by their own nature. These signals are required to be carried from the place in which they are generated to the central processor, which may be located even more than 100 - 200 m away from the transducers that generated the information. Along this path, signals are mixed with those that come from the other transducers in the same structure and connected on the same cable. Furthermore, noise is captured from the whole structured and not from the individual transducer of interest. In addition, a high-impedance transducing system may be affected by considerable electric noise.

For this reason, it is difficult and in certain cases impossible to locate the point where a break-in or a break-in attempt occurred, which will cause false alarms, that users do not easily tolerate.

The systems that rely on sensors with inertial mass transducers can produce much larger electric signals than piezoelectric sensors. In these systems, vibrations cause electric contacts to open, and hence power cut-offs of varying durations in transducer-connecting conductors.

Nevertheless, here information is degraded in that the duration of power cut-offs is only qualitatively related to vibration intensity, and not easily determined in quantitative terms.

Here again, it is difficult and in certain cases impossible to locate the intrusion point, i.e. where a break-in or a break-in attempt occurred, which will cause false alarms, that users do not easily tolerate.

Both systems are usually sold with a cable having transducers pre-wired at predetermined distances, to avoid field wiring, which would result in very high installation costs, to ensure accuracy. However, pre-wired transducers stiffen the cable/transducer system and complicate transport and handling during installation.

Distributed constant systems are usually based on the piezoelectric effect or the triboelectric effect, which may involve rubbing between an inner conductor and a surrounding sheath. For this reason, these cables involve difficulties in manufacture and especially handling during installation.

Their ends are required to be blocked, by special grease materials, to prevent moisture ingress without hindering movement of the inner conductor/s. Furthermore, for effective adhesion to the structure, the cable shall be fixed to the fence with a number of cable ties, which shall ensure adhesion of the cable to the fence without crushing it.

This is a very laborious operation, whose outcome hardly has the required uniformity. These cables are very difficult to handle during installation, and any damage thereto prevents successful installation.

The signals so generated, especially in piezoelectric cables, have a low intensity and shall be carried to the central processor with the noise of the entire structure. Even when complex time-domain reflectometry is used, the intrusion point is not easily located, due to the drawback that is especially caused by the instantaneous and highly variable distance (e.g. from 2 to 200 m) between the intrusion event and the processor, which strongly restricts resolution and accuracy of the system.

This restriction affects any analysis relying on event locating processes. Also, the noise caused by natural events such as rain or wind, adds further restrictions to the location of the intrusion point, and multiple simultaneous intrusions may act as mutual screens.

It shall be further noted that the event of removing the cable from the structure, which is an easy operation, is not even detected by the above described systems, although it would make the system totally inefficient.

Furthermore, all prior art systems cannot ascertain whether the break-in event is caused by a cut on the fence, or by climbing or upward or downward bending of the fence, as analysis simply consists in the comparison of the detected and amplified signal with a level threshold.

With this analysis, the signals caused by events such as rain and wind are almost undistinguishable from other signals caused by fence cutting. If the sensitivity of the system is adjusted to allow cutting event detection, rain and wind would cause a great number of false alarms, which would affect reliability of the information generated by the system.

Technical problem

Therefore, in the field of security systems for monitoring an intrusion detection barrier, there is an apparent need for identification of break-in attempts along the perimeter of the barrier and the position in which such break-in attempt occurred, allowing generation of a reliable alarm signal, free of false positives. Thus, the present invention is based on the problem of providing a security system for monitoring an intrusion detection barrier, that has such functional features as to fulfill the above need, while obviating the above prior art drawbacks.

Technical solution

This problem is solved by a security system for monitoring an intrusion detection barrier as defined in claim 1.

Advantageous effects

The present invention provides a security system for monitoring an intrusion detection barrier that can determine with higher accuracy than prior art systems whether the signals detected by sensors should be attributed to noise such as rain, wind, hail, or are generated by a human break-in attempt, and that can discriminate whether such attempt has been made by climbing, cutting or displacing the fence.

The inventive system is based on a network of MEMS sensors, which are interconnected by subnetworks and joined into a single network all along the perimeter to be monitored. Each subnetwork is monitored by a peripheral analyzer, which can both sample vibration data and perform certain pre-analysis tasks, possibly including comparative analyses between the various channels for a first estimate of the infraction probability, which is preferably performed by fuzzy logics. Then, the data so obtained is transmitted to a central analyzer for comparative analysis of all signals, preferably performed by fuzzy logics.

Also, using the present invention and the analysis capability of each sensor of the system, the response of the latter can be adapted to any heterogeneity of the fences on which the sensors are installed.

Furthermore, the use of MEMS sensors in the intrusion detection system affords highly accurate, repeatable and stable measurements both with time and through different ambient conditions.

Finally, the use of MEMS sensors in the intrusion detection systems allows generation of Volt signals that are not affected by the noise of the whole structure but only of a small part of it, i.e. the part around the MEMS sensor. This affords improved signal-to- noise ratio and improved analysis quality, and reduces or even eliminates false positives.

Brief description of the drawings

Further features and advantages of the method of the present invention will result from the following description of one preferred embodiment thereof, which is given by way of illustration and without limitation with reference to the accompanying figures, in which:

- Figure 1A is layout of an intrusion detection barrier with which the system of the present invention is associated;

- Figure IB is a sectional view of a possible association of the security system with the intrusion detection barrier, according to the present invention;

- Figure 2 shows a single unique assembly of sensors that are part of the intrusion detection system of the present invention;

- Figure 3 shows a plurality of unique sets as shown in Figure 1, such plurality of unique sets being interconnected according to the present invention;

- Figure 4 is a graphical representation of a function indicative of the power that can be associated with a barrier break-in event, such as cutting, according to the present invention; - Figure 5 is a graphical representation of a function indicative of the duration that can be associated with the break- in event of Figure 4, according to the present invention;

- Figure 6 is a graphical representation of the duration of an analysis window for the duration-indicative function of Figure 4.

Detailed description

Referring to the accompanying figures, numeral 1 designates a security system for monitoring an intrusion detection barrier 2 to detect mechanical actions exerted thereon. It should be noted that the term mechanical action on the barrier is intended to indicate actions that can apply a force on the barrier. These mechanical actions include actions that can be exerted on the barrier by man, such as cutting, climbing, lifting, displacing, bitting, etc., and actions that can be exerted by weather agents, such as wind, heavy hitting rain, hail, as well as actions that can be exerted by animals.

Particularly, referring to Figure 1A, the system 1 appears to be associated with the intrusion detection barrier 2 which comprises, for instance, a metal fence or a perimeter wall 1A and/or a glazing, gates or doors 2B but might also comprise armored cabinets (such as bank vaults), safes, safe-deposit boxes or the like, to define a perimeter that delimits the area to be protected.

The security system 1 comprises a plurality of sensors M_j with Kj<K, which are connected in signal communication, e.g. by a special communication cable and are at least partially arranged, preferably along the entire perimeter of the intrusion detection barrier 2. Particularly, the plurality of sensors M_j is composed of Micro Electro-Mechanical Systems sensors (in short MEMS), which are adapted to generate at least one electric signal V_ei in response to a mechanical action on the intrusion detection barrier. For example, the use of MEMS sensors in the system 1 requires such MEMS to be placed in a housing 10 that can ensure one or more of the following properties: tightness, weather resistance, electromagnetic noise shielding and isolation from the barrier to be monitored.

For this purpose, for example the housing 10 will be made from a thermoplastic polymer material, such as polycarbonate.

Furthermore, also referring to Figure IB, the housing will comprise a bottom 10A, a support base 10B for the MEMS M_j, from which a closing lid 10C extends to enclose such MEMS, the support base 10B being associated with the bottom 10A by fastener means 10D. The housing 10 is designed to be associated with the barrier 2 such that the barrier 2 can be interposed between the bottom 10A and the support base 10B. By this arrangement, the MEMS M_j will be located in the proximity of the barrier 2 to detect any vibration therein, and the conformation of the housing 10 will prevent any oscillation due to its structure from distorting measurements.

In a preferred embodiment, MEMS sensors are uniaxial, biaxial or triaxial sensors and can generate an electric signal V_ei of the order of a few Volts, e.g. three Volts, along each of their axes. Therefore, this value is about three orders of magnitude larger than other prior art sensors, of either lumped or distributed constant type. Furthermore, the signal V_ei so generated is not affected by the noise of the entire structure, but only of a small part of it, i.e. the part around the MEMS signal that generated the signal Therefore, the signal V_ei is not affected by the noise of the whole barrier, and its individual, local treatment improves the signal-to-noise ratio and hence the quality of the next analyses. For example, this signal V_ei is indicative of the power or other synthesis parameters (determined over appropriate intervals of time) of the mechanical action exerted on the intrusion detection barrier.

It shall be noted that, according to a peculiar aspect of the present invention, once the MEMS sensors M_j have been associated with the barrier, each of such MEMS sensors Mj defines a portion or zone 2A of the perimeter of the barrier 2.

Particularly, each of these MEMS sensors M_j defines less than 1% of the overall perimeter of such barrier, preferably 0.5% of the perimeter of the barrier 2.

In an advantageous aspect of the present invention, the plurality of MEMS sensors M_j are divided into a plurality of unique sets SBSj with 0<i<N, where N is for instance seventy, and where each unique set SBSi comprises at least one MEMS sensor M_j, thereby forming a plurality of subnetworks.

In other words, the MEMS sensors M_j are organized into a modular structure (each module consisting of a unique set), so that multiple modules of MEMS sensors Mj form the intrusion detection system of the present invention.

The sensors of the unique set are spaced at a regular pitch P, e.g. five meters, but they might also be arranged with an irregular pitch P.

In one embodiment, each unique set SBSj is preferably composed of an odd number of sensors. In a preferred embodiment, the unique set comprises seven MEMS sensors M_l .., M₇, as shown in Figure 1A concerning the unique set SBSi₌₁, although a smaller number of MEMS sensors may be also provided like, for instance in the unique sets

which are composed of three, two and one MEMS sensors respectively.

It shall be noted that each unique set SBSj defines a zone 2B of the barrier 2, such zone 2B being larger than the portion 2 A defined by an individual MEMS sensor M_j.

The intrusion detection system 1 comprises a microcontroller 3, which is in signal communication with the plurality of MEMS sensors Mj.

Particularly, the microcontroller 3 is configured to respond to each of the (one or more) electric signals V_ei generated by each sensor M_j.

As described below, the microcontroller 3 is configured to receive and process one or more electric signals V_ei to generate an alarm signal Vaii representative of the portion 2A and hence the zone 2B of the barrier 2 in which the mechanical action was exerted.

The microprocessor 3 is configured to identify the type of such mechanical action, which means that it is configured to identify which zone 2B of the barrier 2 has experienced the mechanical action, according to the reception of the signal V_ei generated by one or more MEMS sensors Mj of that unique set or multiple unique sets.

In order to notify users where the mechanical action actually took place on the intrusion detection barrier, the system 1 comprises a communication device 4 adapted to communicate such alarm signal V_an (see Figure 3).

For instance, the communication device 4 is in the form of an electrical/electronic apparatus that can emit audible sounds, signal lights, communications displayed on a screen, communications transmitted via GSM or the like, etc.

The microcontroller 3, also referring to Figure 2, comprises a peripheral analyzer AP, and particularly each unique set SBSj comprises the peripheral analyzer AP, so that each MEMS sensor M_j of a particular unique set SBSi is in signal communication with such peripheral analyzer AP via its communication channel.

In other words, a peripheral analyzer AP (e.g. in the form of a microcontroller) is provided for each unique set SBSi, and is in signal communication with all the MEMS sensors M_j that form such unique set SBSj, and is configured to generate an alarm signal Sail ' , as a function of the electric signal/s V_ei generated by the MEMS sensors of each unique set SBSj.

It shall be noted that the peripheral analyzer AP of each unique set SBSj may coincide with one of the MEMS sensors Mj.

Thus, also referring to Figure 2, the MEMS sensor j=4 coincides with the peripheral analyzer AP. Here, all the remaining MEMS sensors of the unique set communicate with such peripheral analyzer AP, i.e. with the microcontroller via a particular communication channel.

The number of MEMS sensors M_j in each unique set is selected according to the maximum number of MEMS that a peripheral analyzer AP can handle with its memory capacity, considering that the microcontroller of the AP shall have a small size, to avoid excessive power consumption, and hence shall have memory restrictions. It should be also noted that in the preferred embodiment of the unique set (i.e. the set composed of seven MEMS sensors), by processing the signals from seven MEMS spaced at a pitch P of five meters, the AP may make inferences about events, such as rain, of a rather large but not too large set. It shall be finally noted that the use of an odd number is advantageous because the peripheral analyzer AP itself contains a MEMS and hence, the connecting structure has to be balanced left and right by providing the same number of sensors both to the right and the left of the AP. This will provide a repeated structure for unique sets SBSj, equal to the previous and the next ones.

Furthermore, each peripheral analyzer AP of each unique set SBSj comprises a mass storage medium, for storing the coordinates that define the distance between each pair of adjacent MEMS sensors M_j of the unique set. The origin is arbitrary and may coincide with the position of the MEMS sensor M_j=1.

In other words, the storage of the peripheral analyzer contains mutual distances for each MEMS of the unique set, which means that each AP stores the positions of the MEMS sensors of the unique sensor, and particularly the distance between the sensors of the unique set.

Each peripheral analyzer AP comprises firmware configured to:

- sample the signals V_ei for the vibration data detected by the MEMS sensors Mj of that particular unique set;

- analyze such signals V_ei to provide a first estimate of the break-in probability on the barrier 2, such analysis being implemented with Boolean logics and/or preferably fuzzy logics.

For this purpose, the peripheral analyzer AP of each unique set SBSj is adapted to detect, preferably continuously, one, more or all electric signals V_ei generated by each. MEMS sensor Mj of the unique set SBSj; comparing the electric signal/s V_ei so detected with a plurality of reference signals S_ref, the latter being previously stored in such mass storage medium, and generating the alarm signal Van according to the result of such comparison, i.e. when the electric signals V_ei , e.g. their modulus values, are greater than one of the plurality of reference signals S_ref during an analysis interval Tanai- It shall be noted that, if a Boolean logic is used, the reference signals S_ref will be the thresholds that uniquely delimit the inclusion of the instantaneous signal (e.g. the power) within the set of break-in signals or the set of non-break-in signals on the barrier. If a fuzzy logic is used, the reference signals are not that clear-cut, but their blurred character affords more accurate determinations. This is because a signal V_ei is no longer compared to a single threshold S_ref, but it is as if it were compared to hundreds of different thresholds and, if it exceeds them, it will belong time after time to different sets of signals, i.e. having characteristics differing, possibly to a small extent, from those of other signals, which affords much more accurate analysis of the signal generated by the sensors and better understanding if its origin (i.e. whether it was triggered by a break-in attempt or noise).

It shall be noted that the firmware of the peripheral analyzer AP is configured to generate such alarm signal V_au according to the space coordinates of the MEMS sensors of the particular unique set. Therefore, any mechanical action exerted on the barrier is recognized and preferably also correlated with the MEMS sensors adjacent to the one that generated the signal/s, so that the position of such mechanical action on the barrier can be precisely identified.

Therefore, the microprocessor 3 is configured to generate an alarm signal S^i in response to the break-in event, such alarm signal Saii being indicative of the portion 2A (and/or the zone 2B) of the barrier 2 that experienced the mechanical action.

An analysis interval Tanai is intended herein as the time during which a modulus of the signal V_ei (e.g. the intensity modulus) is equal to or greater than that of the reference signal/s

Sref, The characteristics of the firmware of the peripheral analyzer AP included in every unique set SBSj allow the response of the system 1 to bg adapted to any type of intrusion detection barrier.

Advantageously, since each MEMS sensor is distinct from all the other sensors both in terms of detection and analysis, and since each MEMS sensor is responsible for its portion 2A of the whole barrier, any electric signal V_ei resulting from other simultaneous mechanical actions will have no influence on the other MEMS sensors, whereas with other techniques and architectures they might interfere with and mask one another.

The firmware of the peripheral analyzer AP of the unique set SBSj is optionally configured to generate the alarm signal Van based on the comparison between the electric signal/s V_ei generated by the plurality of MEMS sensors of such unique set SBS,.

The above described analysis for each MEMS sensor M_j, possibly repeated for a series of adjacent sensors, allows validation and invalidation of the recognition of an event instead of another, e.g. fence cutting instead of rain or other noisy events.

Here, the firmware of the peripheral analyzer AP is adapted to preferably process the signals V_ei received for each of the three axes of the MEMS sensor, i.e. it processes three acceleration signals V_ei that each sensor MEMS generates, and measures, preferably in a continuous manner, the signals V_ei that each MEMS M_j detects. If the signals V_ei have a higher intensity than the reference value at the start of the analysis (e.g. the reference value of the mechanical action that corresponds to a cut of the barrier), then the time in which the power value (or another selected parameter) is above said threshold during the analysis window Tanai- These two values (i.e. |V_ei|>|S_ref| and Tanai) are each screened through multiple membership functions, so that fence cut signals may be determined and distinguished from the others, which do not have the same or a similar behavior in their development.

Each of these functions is used to determine to which extent, for instance over a range from zero to one, each value V_ei[>|S_ref| and the duration T^i are attributed to the mechanical action exerted on the intrusion detection barrier.

A value V I of one indicates that a particular mechanical action (e.g. a cutting action) is concerned, whereas a value of zero indicates that the action concerned is certainly caused by another phenomenon.

By adding this value Vj to any other value V_x detected within a maximum time T_maxin the analysis window Tanai to obtain a value V_max, and by comparing the value V_max to the predetennined threshold of pulses that will be certainly associated with that special type of mechanical action, here the cutting action (i.e. with a membership value 1), false alarm rate may be improved, without affecting detection probability, as a real breach can be only made in the barrier by multiple (at least five) cuts.

The firmware of the peripheral analyzer AP is configured to check consistency of the minimum time interval T_mi_n between one mechanical action and the next. Also in this case, the membership value ranges from zero to one, and hence the actual membership value of the last pulse is corrected by multiplying it by this new factor. The last added value is subtracted from the previously accumulated value, and is multiplied by the new membership value for this function and added again to the previously accumulated values. If the time from the last pulse reaches a maximum value during Tanai, which is set at each new pulse, the accumulated value and all the timers are reset, which will terminate the analysis without generating and storing anything.

Referring to Figures 4, 5 and 6, there will be now described the method of analysis by the firmware of each MEMS M_j of a particular unique set SBSj, when the mechanical action on the barrier is a cutting action. It shall be noted that a number of instrumental measurements were taken by the Applicant for detecting intrusion events by cutting (and others) on various types of fences. Therefore, for the most common fence types, characteristics were isolated that distinguish these events from those that characterize natural events such as wind and rain and events caused by roads or railways near the protected fence, as well as noise generated by small animals that may accidentally hit the fence. Particularly, characteristics were detected for events of possibly minor intensity (i.e. "fingerprints" or "signatures") that need to be notified, such as fence cutting.

Referring now to Figure 4, a membership function 11 is described for measured barrier cut power. Since barrier cutting events are characterized by relatively short pulses, within the universe of possible durations of detected power peaks, ranging from 0.2 to 5 sec, a duration membership function 12 is described, with reference to Figure 5, to define inclusion in the fuzzy set of fence cut signal power durations. In view of the fact that cutting events must be separated by a given minimum period of time to be considered as such, and that they should not be accounted for if nothing happens for more than a given time, in the universe of intervals from one cut to the other, ranging from zero to 80 seconds, a membership function 13 is described, with reference to Figure 6, for inclusion within minimum and maximum time intervals between two fence cutting actions. Considering that a breach on the barrier requires a considerable number of cutting actions, e.g. at least four or five, if each detected cutting action has a quantitative value equal to the product of values obtained for each of the above described membership functions, then by adding up such results and comparing them with those expected (set), the detection of a fence cutting attempt is highly reliable, both in terms of detection probability (DP) and Nuisance Alarm Ratio (NAR). If the time between two successive events exceeds the maximum time interval (here 75 seconds), the count of accumulated events with their values is reset.

The signals received from each MEMS sensor Mj of the particular unique set SBS; for barrier cutting events are characterized by intensity, duration and repetitiveness factors, which are analyzed in a very localized manner. This will provide an additional degree of validity, using a fuzzy analysis of power peak intensity distribution, as detected by the various MEMS sensors adjacent to the one that generated the signal/s V_ei. of a single unique set SBSj. Assuming the highest intensity value of the signal V_ei (e.g. representative of the maximum power peak) generated by a particular MEMS sensor the peaks of the signals detected by the adjacent (previous and next) MEMS sensors may only have a lower intensity, or an equal intensity may be only detected on the previous sensor or on the next sensor. The peak of signals V_ei detected by further previous or next sensors may only have a lower intensity. The operators equal, less than and greater than are fuzzy operators, and are defined by membership functions. If this distribution does not occur, a pulse with all the other characteristics equal to a cutting pulse will not be deemed as valid. Indeed, when moving away from the MEMS sensor that generated and detected the maximum peak of the signal V_ei, the peaks generated and detected by the other MEMS sensors can only decrease.

It shall be noted that the firmware of each MEMS or the firmware of the peripheral analyzer AP are configured to store a value Vgj- representative of the gravity acceleration experienced by each MEMS sensor and to generate an alarm signal whenever such value Vgr is changed from the stored value.

For instance, the alarm signal V_an is generated if the value Vgr is changed by a number of degrees exceeding a predetermined amount and anyway variable (e.g. 5°). In other words, the ability of the MEMS sensor to sense the gravity acceleration experienced by the MEMS sensor, to store such value in the storage of the peripheral analyzer AP (e.g. during installation of MEMS sensors on the barrier) and generate the alarm signal if a mechanical action attempt is made, such as dislocation of the MEMS sensor, either due to removal thereof from the surface of the barrier to be protected or to bending of the barrier to open a breach therethrough. MEMS sensors are configured to provide a signal proportional to gravity acceleration and the measurement of gravity acceleration on one or more axes of the accelerometer provides information about the position of the accelerometer in space.

Referring now to Figure 2, in the system 1 the plurality of unique sensors SBSj are interconnected. For this purpose, the microcontroller 3 comprises a central analyzer AC which is advantageously in signal communication with each unique set SBSj.

For example, such signal communication is provided as a warning through the actuation of one or more relays, or a notification of the event, with the sector and distance data, through an RS485 serial communication network or an Ethernet network and/or other types of remote connection). The central analyzer AC has its own mass storage medium, . which can store all significant signals generated by said plurality of MEMS sensors M_j, as well as the distance coordinates of all the MEMS M_j in the various unique sets SBSj.

It shall be noted that each peripheral analyzer AP of a particular unique set SBSj generates an alarm signal Vaii' , an that the central analyzer AC is configured, by firmware, to obtain a comparative analysis of all signals, preferably using fuzzy logics.

For this purpose, the central analyzer AC is configured to detect one or more alarm signals V_an' generated by the peripheral analyzers AP of a given unique set SBSj; process the alarm signals V_an' generated by adjacent (previous or next) unique sets SBSi; generate the alarm signal Vaii as a result of such processing, to notify whether two or more zones 2B (i.e. two or more unique sets SBSj) have detected the same or different mechanical actions on the intrusion detection barrier.

In other words, the central analyzer AC both performs correlations that cannot be made by the individual peripheral analyzers AP, and allows simultaneous detection of mechanical actions on two or more distinct zones 2B of the barrier 2, such zones being either adjacent to (e.g. when the mechanical action is exerted at the boundary of two zones, i.e. of two unique sets) or remote from each other (e.g. when the mechanical action is exerted on two distant, non-contiguous zones).

Those skilled in the art will obviously appreciate that a number of variants may be envisaged to the above intrusion detection system, still within the scope of the invention, as defined in the following claims.

Claims

1. A monitoring system for monitoring an intrusion detection barrier (2), said monitoring system comprising a plurality of sensors (M_j) arranged along at least part of the perimeter of said intrusion detection barrier and designed to be firmly attached thereto, and a microcontroller (3, AP, AC) in signal communication with said plurality of sensors, said security system being characterized in that:

- said plurality of sensors (M_j) are MEMS sensors and are divided into a plurality of unique sets (SBSi), each unique set comprising at least one MEMS sensor and defining a zone (2B) of said intrusion detection barrier, each MEMS sensor generating at least one electric signal (Vei) in response to one or more mechanical actions on said intrusion detection barrier (2);

- said microcontroller (3, AP, AC) being configured to respond to each of said electric signals (V_ei) generated by said plurality of MEMS sensors, said microcontroller (3, AP, AC) being configured to detect said mechanical action when said at least one electric signal (V_ei) is received, said microcontroller being configured to identify said zone (2B) of said barrier based on the reception of said at least one electric signal (V_ei) corresponding to one of said plurality of said unique sets of MEMS sensors, said microcontroller being configured to generate an alarm signal (Vaii) in response to said mechanical action, said alarm signal being indicative of said at least one zone (2B).

2. A monitoring system as claimed in claim 1, wherein said microcontroller (3, AP, AC) comprises a peripheral analyzer (AP), each unique set comprising said peripheral analyzer, each MEMS sensor of the relevant unique set being in signal communication with said peripheral analyzer via its communication channel, the analyzer receiving said at least one electric signal (V_ei) and generating said alarm signal (V aii) based on one or more fuzzy logic rules.

3. A monitoring system as claimed in claim 1, wherein said microcontroller (3, AP, AC) comprises a peripheral analyzer (AP), each unique set comprising said peripheral analyzer, each MEMS sensor of the relevant unique set being in signal communication with said peripheral analyzer via its communication channel, the analyzer receiving said at least one electric signal (V_ei) and generating said alarm signal (V_aii) based on a Boolean logic.

4. A monitoring system as claimed in claim 2 or 3, wherein each peripheral analyzer (AP) of each unique set comprises a first mass storage medium, for storing the distance coordinates (x) of each MEMS of the unique set.

5. A monitoring system as claimed in any preceding claim from 1 to 4, wherein each peripheral analyzer (AP)^' of each unique set comprises first firmware, which is configured to:

- continuously detect said at least one electric signal (V_ei) from a MEMS sensor of its unique set;

- compare said at least one electric signal/s (V_ei) with a plurality of reference signals, the latter being previously stored in such mass storage medium;

- generate said alarm signal (Vaii) according to said comparison when said at least one electric signal (V_ej) is greater than one of said plurality of reference signals during a predetermined time interval.

6. A monitoring system as claimed in claim 5, wherein said first firmware is configured to generate said alarm signal as a function of said distance coordinates (x) for said MEMS sensor of the unique set.

7. A monitoring system as claimed in claim 5 or 6, wherein said first firmware is configured to generate said alarm signal based on the comparison between said at least one electric signal (V_ei) generated by said plurality of MEMS sensors of said unique set.

8. A monitoring system as claimed in any preceding claim, wherein said microcontroller comprises a central analyzer (AC), said plurality of unique sets being interconnected to one another and to said central analyzer.

9. A monitoring system as claimed in claim 8, wherein said central analyzer comprises a second mass storage medium for storing all significant signals generated by said plurality of MEMS sensors.

10. A monitoring system as claimed in any preceding claim from 8 to 9, wherein said central analyzer (AC) comprises second firmware, which is configured to:

- detect one or more alarm signals (V_aii' ) generated by the peripheral analyzers (AP) of a particular unique set (SBS;);

- process said alarm signals (V_an' ) generated by adjacent unique sets (SBSj);

- generate said alarm signal (Van) as a result of such processing, to notify whether two or more zones (2B) have detected the same or different mechanical actions on said intrusion detection barrier.

11. A monitoring system as claimed in any preceding claim, wherein each MEMS sensor of said plurality of sensors is a uniaxial, biaxial or triaxial sensor, said MEMS sensor being able to generate said electric signal along one axis, two axes or three axes respectively.

12. A monitoring system as claimed in claim 4, and in any claim from 10 to 11, wherein said first firmware of each MEMS sensor is configured to store in said mass storage medium a value representative of the gravity acceleration experienced by said MEMS sensor, said value being detected along an axis of said MEMS sensor.

13. A monitoring system as claimed in any preceding claim, wherein each unique set comprises an odd number of MEMS sensors.

14. A monitoring system as claimed in any preceding claims, wherein each MEMS sensor (M_j) is placed inside a housing (10), comprising a bottom (10A), a support base (10B) for supporting said MEMS (M_j), said support base (10B) being associated with said bottom (10A) by fastener means (10D), said intrusion detection barrier (2) being interposed between said bottom (10A) and said support base (10B).

15. A monitoring system as claimed in any preceding claim, comprising a communication device (4) which is adapted to communicate said alarm signal (Van).