Principles for Intrusion Detection
        A security system is only as good as its weakest link
        
        
			- By Pete Accetturo
- Feb 01, 2012
The best defense is a good offense. That often-used sports principle
  applies to military applications. It is an axiom that applies to the security
  industry, as well. When it comes to perimeter security, though,
  a strong offense provides a good defense, but that defense will be
  only as strong as its weakest link. The question to be answered is how
  good is that defense? Or where is the weakest link?
  
If there are weak links, then you have unprotected perimeters, which means unprotected
  assets, unprotected people and, inevitably, security breaches. The ramifications
  of these breaches can be disastrous, so the threat of intrusion remains a
  prime concern at all facilities. Because many perimeters are simply too long for
  conventional security patrols to cover practically or effectively, advanced perimeter
  intrusion detection systems (PIDS) have become the only answer.
  
Historically, PIDS have had some trouble clarifying between a true positive and
  a false positive (nuisance alarms). All too few PIDS provide sufficient analytical
  tracking, real-time assessment or any environmental capability, making it extremely
  difficult for security officers to respond to the ingress or egress point in a prompt
  manner. Technology has advanced to the point of offering a wide array of sensing
  solutions for PIDS. Those PID solutions come with an array of efficiency, costeffectiveness
  and precision. There are some important factors to consider when
  reviewing current PID technologies:
  -  System durability/reliability;
- Minimal nuisance alarms (false positives)
- Maximum detection capability;
- Minimal maintenance
- Ability to accurately pinpoint the location of intrusion; and
- Ability to work with other/complementary technologies.
Since the dawn of civilization, maintaining perimeter security has been a top
  priority. Having some form of it—whether a city wall, a moat or even trained
  animals—has always been critical for stopping, delaying or deterring a perimeter
  breach. Even with primitive perimeter security systems, having real-time intelligence
  and immediate response is vital. The difference between being aware of a
  breach and the intelligence to properly respond can truly mean life or death.
  
While technology has advanced over the years, the security fundamentals have
  not. Today we have both active and passive intrusion detection technology. Video
  analytics; sonic, infrared, radar and fiber optic sensing; and buried vibration and
  sonic or attenuation detection all have significant roles in PIDS. Fundamentally, good physical security is a combination
  of four defensive principles: deterrence,
  delay, detection and denying a breach.
  The first two actions are considered
  passive defense while the latter two are
  active in nature.
  
Deterrence
  
The goal of physical security is to convince
  potential attackers that the likely
  costs of attack exceed the value of making
  the attack (e.g., that consequences
  of a failed attack may well exceed the
  gain). The combination of layered security
  features establishes the presence
  of a security-rich deterrence system.
  
The initial layer of security for
  a campus, building, office, or other
  physical space uses crime prevention
  through environmental design to deter
  threats. Some of the most common
  examples are also the most basic—
  warning signs, fences, vehicle barriers,
  vehicle height-restrictors, restricted
  access points, site lighting and trenches.
  However, even passive things such
  as hedgerows may be sufficient in some
  circumstances.
  
Delay
The next layer is mechanical and includes
  gates, doors and locks. Key control
  of the locks becomes a problem
  with large user populations and any
  user turnover. Keys quickly become unmanageable,
  often forcing the adoption
  of electronic access control. Electronic
  access control easily manages large user
  populations, controlling for user lifecycle
  times and dates and individual
  access points.
  
For example a user’s access rights
  could allow access from 7 a.m. to 7
  p.m., Monday through Friday, and
  expire in 90 days. Another form of access
  control includes the use of policies,
  processes and procedures to manage
  ingress into the restricted area. The deployment
  of security staff conducting
  checks for authorized entry at predetermined
  points of entry is usually supplemented
  by mechanical and electronic
  access control or simple devices such as
  physical passes, ID and RFIDs.
  
Detection
  
The third layer is intrusion detection
  systems or alarms. Intrusion detection
  monitors for unauthorized access. It is
  less a preventative measure and more
  of a response trigger, yet it has a high
  incidence of false alarms.
  
In many jurisdictions, law enforcement
  will not respond to alarms from
  intrusion detection systems. For example,
  a motion sensor near a door could
  trigger on either a person or a squirrel.
  A simple sensor does not do identification,
  and as far as it is designed, anything
  moving near that door is unauthorized.
  That is why CCTVs are the
  next step in a layered security system
  approach. Identification of the intrusion
  event is essential for the required
  intelligence detection.
  
Security cameras can be a deterrent in
  many cases, but their real power comes
  from incident verification and historical
  analysis. For example, if alarms are being
  generated and there is a camera in place, the camera could be viewed to verify the alarms. In instances when an attack
  has already occurred and a camera is in place at the point of attack, the recorded
  video can be reviewed.
  
Denying a breach
  
Guards have a role in all layers: in the first, as patrols and at checkpoints; in the
  second, to administer electronic access control; in the third, to respond to alarms.
  The response force must be able to arrive on site in less time than it is expected the
  attacker will be required to breach the barriers.
  
In the fourth layer of perimeter security, a guard’s role is to monitor and analyze
  video. Users obviously have a role also by questioning and reporting suspicious
  people and aiding in identifying people as known versus unknown. Often,
  photo ID badges are used and are frequently coupled to the electronic access control
  system. Visitors are often required to wear a visitor badge.
  Every application is unique, depending on the type of facility being protected,
  its operating environment, its perimeter fence construction, its intrusion and security
  history and the perception of threats.
  
Each individual site has a unique set of parameters that require a customized
  perimeter solution. Just as the military cannot apply a one-solution strategy for all
  military conflicts, neither can any perimeter apply a one-size-fits-all solution strategy.
  Goals need to be established that are influenced by environmental concerns
  (including any severe topography factors), the flow of traffic (foot and auto) and
  the possibility of any major weather concerns. Effective PIDS account for all of
  these elements and take budget into consideration. These factors weight the PID
  system in varying degrees that impact its overall efficiency. Again, a layered system
  is the best solution.
  
Once a PIDS solution is installed, the next critical step will involve a complete
  calibration of the system. This calibration must account for all environments, sensitivity
  and duration of any possible event breach. This ensures that any true positive
  alarms will provide the best intelligence for the appropriate response. Thus a
  proper PIDS solution provides the necessary forewarning for deterrence, delay,
  detection and denying a breach.
  
Important design considerations derived from the aforementioned factors produce
  a well-integrated PIDS. Design elements to be considered are:
  Environmental design—Having intrusion detection sensors installed in the field
  or on the fence.
  
Mechanical, electronic and procedural access control—Having an alarm processor
  that drives and analyzes the raw sensor signals.
  Intrusion detection, with appropriate response procedures—Having a security or
  alarm management system that notifies security staff of an alarm and the location
  of the intrusion.
  
Personnel identification—Having a communications infrastructure that ties
  these three elements together and connects the system to the security staff,
  along with an established and clearly documented site policy and alarm response
  procedure.
  
Militarily or commercially, whatever the perimeter’s greatest strength is, if left
  ungaurded, it becomes its greatest weakness. Strengths and weaknesses of a security
  system require a clear definition so policies and procedures can ensure appropriate
  action.
  
Intrusion detection technology will continue to advance, but
  these advances will be focused more around software than hardware
  as manufacturers continue to pursue improved system performance,
  flexibility, and reliability.
        
        
        
        
        
        
        
        
        
        
        
        
        This article originally appeared in the February 2012 issue of Security Today.