Software Products
IrisAccelerator™
Fast Iris Matching System
IrisAccelerator is a fast iris matching system built on Iris ID's technology, configurable to support large population databases of up to 1 billion with hundreds of millions of de-duplication transactions per day.
Overview
IrisAccelerator is a system that provides a fast iris matching facility using Iris ID's technology. IrisAccelerator is configurable to support large population databases of up to 1 billion with hundreds of millions of de-duplication transactions per day. The IrisAccelerator system is a cluster of servers where each server runs multiple matching engines, often referred to as 'Blades'. The IrisAccelerator architecture is designed to be hardware independent and can be deployed on a virtual machine to realize economies of scale. In order to reduce the response time of searching the database, IrisAccelerator™ matches an enrollee's IrisCode in a parallel fashion.
The IrisAccelerator™ system consists of the following modules. The IrisAccelerator™ SDK provides the Application Program Interface to communicate with the system, with enrollment and identification functions that can be used by end-user applications. IrisController™ is the server component that manages the database of users and their IrisCodes, processing SDK requests as the one-point access to IriSABRE™. IriSABRE™ is the iris matching module that matches requests against the entire database; all enrollments and IrisCode deletions are updated in IriSABRE™. The LoadBalancer runs on a Linux server and interfaces between the SDK and the cluster of groups, forwarding requests to IrisControllers in round-robin style, adding new servers to the group and removing non-functioning ones, with a failover protection mechanism that ensures smooth functionality of the cluster. The IrisAccelerator™ Configurator provides an interface for creating, deleting, and modifying the groups and servers in the system. The IrisAccelerator™ Device Manager ('DM') provides control and monitoring of the entire system, displaying the configuration of every server as well as the active and backup LoadBalancer.
Multiple deployment configurations are available, from an enterprise blade-server platform to private cloud solutions.
Key Features
Scalability
Scales beyond 50 million records by running multiple matching engines in parallel across groups.
High Availability
No single point of failure across the clustered architecture.
Partial Failure Support
The system continues operating even when individual components experience failures.
Component Recovery
Non-functioning servers are removed from the group and recovered without disrupting the cluster.
Consistency
All enrollments and IrisCode deletions are consistently updated across the matching modules.
Speculative Execution
Speculative execution reduces database search response times through parallel matching.
Specifications
Technical data
- Database Capacity
- Configurable to support large population databases of up to 1 billion identities
- Throughput
- Hundreds of millions of de-duplication transactions per day
- Enterprise/Blade Platform
- Runs on IBM Blade Server platforms; cluster of blade server chassis with multiple blades; one blade runs IrisController to control blades within a chassis; clusters made operational by LoadBalancer; Oracle database (RAC/standalone) for data storage
- Private Cloud Entry Level Solution
- Matching engines on 'pizza box' servers; high throughput and fault tolerant by running matching engines in series; supports up to 250,000 records; upgradeable; logging; Oracle database (RAC/standalone)
- Private Cloud Solution
- Cluster of matching engines on different 'pizza box' servers; hardware-independent 'cloud'-based ABIS engine; scalability (>50M) by running matching engines in parallel; logging (debug, info, error, transaction); SNMP support; Oracle database (RAC/standalone)
Ready to deploy this in your program?
Talk to our solutions team about integration, pilot programs, and volume pricing.
Contact sales