the levl-iq platform:

Device Intelligence

solving the issues associated with mac address randomization

The LEVL-IQ platform is built to provide network operators device intelligence and device identity while still limiting user privacy impact. The platform is an easy to deploy SaaS system that accurately identifies devices in just a couple of seconds. The LEVL-IQ platform requires nothing be put on a user device, dramatically simplifying and accelerating adoption. At the core the system is the the unique device identifier, the LEVL-ID. The LEVL-ID is generated automatically from within the network without using any Personally Identifiable Information (PII) from the user.  You can see the latest hospitality trial data here. 

Key Capabilities

LEVL provides network operators with the device intelligence for each device on the network. It provides key information than enables operators to run an efficient and easy to use network. There are four components to device intelligence:

Device Identity is the fundamental building block of our system. Historically device identity was determined using the device MAC address. However recent changes in user-privacy techniques are rendering it useless. LEVL identifies devices by how they “communicate and behave” on the network. We create a device identity in 30 seconds and can then identify the same device in about 2 seconds.

Device Context tells us what type of device is on the network. Is it a phone or laptop? Who makes it? what is the model type and operating system version. Coupling this with device identity allows network operators to assess what types of devices are on the network. This in turn enables more accurate network planning and deployment of support resources. Learn more about device context here.

Device Experience relates to the quality of the service the device is getting on the network. By knowing the device type you can assess if there is enough bandwidth for the device. Different device types have different needs from the network. We can assess the device experience once we know the device type.

Device Behavior is driven by device type as well. By assessing the actual behavior against the known device type LEVL can identify compromised devices and devices that are being spoofed by something else. 




The Device Identity Package is our entry level platform utilizing basic device identity features.  This allows network operators to create, maintain and match device identities on the network. The system features accuracy of >99.5% and typically identifies devices in less that 2.5 seconds.  Device identities can be stored for up to 30 days. The system comes with basic reporting and analytics capabilities and APIs for quick integration into a networks authentication system. The platform can be deployed via a virtual machine on an existing edge compute device or via a small network appliance. The compute power needed is dependent on property size but has been successfully deployed on Intel ATOM platforms with 2 gigabytes of memory.


The Device Performance Package provides operators with device type ( phone vs laptop vs other), device manufacturer, device model and OS information. This package also allows for longer storage periods for device identities and provides a more advanced reporting and analytics package.


The Device Intelligence Package is our top of the line offering and adds device experience and device behavior analysis capabilities. Device QoS, relative to the device type is analyzed and potential QoS issues can be highlighted. Compromised, rogue and spoofed device detection is also included as is multi-site capability.   



Deployment is designed to be easy and time efficient. The cloud portion of the system is configured first, based on the package purchased and the options selected. On site, a small amount of edge compute capability is needed. It can take the form of a virtual machine or a dedicated network appliance.  A raspberry Pi class device compute capability with 2 gig of RAM is all it takes. Access to the DHCP server via port mirroring and a couple of Ethernet ports round out the requirements.  You can read about our ongoing successful deployment in the hospitality industry here.


The specifications, performance and requirements of the LEVL-IQ platform are detailed in the following table:
Devices supported Windows, Mac, Linux, Embedded (IoT), iOS, Android
Device connection types supported Wireless (Wi-Fi, including 802.11ax), and wired (Ethernet up to 1Gb/s speeds).
Configurations supported
  • Single AP (bridge or router mode)
  • Wi-Fi Mesh
  • Multi AP (using switch port mirroring)
Requirements from devices None
User involvement None (100% zero user touch)
Apps, certificates installed on devices None
Average time to identification Less than 2.5s
True identification accuracy More than 99.5% (based on more than 2M connections verified across more than 30K unique devices)
Operating system requirements None
  • Device identification (LEVL-ID)
  • Device context (type, make, model, manufacturer)
  • Device experience (QoS, BW quality) - in roadmap
  • Device behavior (anomaly detection) - in roadmap
LEVL-ID retention period 30 days
Average upload BW requirements 20 KB per connection event
Virtual/Physical appliance specification
  • 2 vCPU
  • 2 GB RAM
  • 20 GB Storage
  • 2 NIC Ethernet
APIs available Supporting both a “push” (asynchronous notification via HTTP callback, Kafka, etc.) and “pull” (query via REST API) model for providing device identification results
Logs available Detailed events at device level for each connecting device across all networks including the device identification and device context.
Statistics available Comprehensive analytics on network unique devices sliced by device type, operating system, device model and more.

In addition, the LEVL-IQ platform offers the following unique capabilities which are only available from LEVL:

Full-Stack device intelligence: The LEVL-IQ platform is capable of leveraging data from all seven layers of the communication stack, including the Physical Layer for device identification. Being able to process and leverage the Physical Layer (when available) future proofs LEVL’s solution as the Physical Layer information is not affected by any future software updates. This capability is only available for AP-based deployments or in scenarios where the customer has the Physical Layer information available at their backend. One such example is if the physical layer data is already processed by the customer for Wi-Fi presence detection capabilities.
Realtime performance: The LEVL-IQ platform is the only available solution which can identify devices in real time (less than 2.5 seconds on average for device identification) while other existing solutions can take up to a few minutes.

State-of-the-art accuracy: LEVL-IQ’s more than 99.5% accuracy is state-of-the-art and is not comparable to anything in the industry.

DOwnload the LEVL-IQ Product Brief

Learn more about LEVL-IQ by downloading our product brief.

find your answer to device identity and mac randomization issues

LEVL helps you maintain and enhance network services while making it easier to manage user privacy. Reach out to learn more.