Security mitigation objects.

ELIXIR Beacon Network Security

Object Design

Introduction

In order to effectively model an attack across an open beacon network, objects will be required that model the source of the attack and the steps taken to mitigate it at the bare minimum..

There are two types of attack possible against a Beacon Network (see:user Requirements Document). A (D)Dos attack which attempts to shutdown a beacon with disruptive queries or a re-identification attack where a malicious user attempts to identify an individual within a beacon.

How these attacks are identified by individual beacons is beyond the scope of this document. However it is also possible for a re-identification attack to be packaged within a DDos attack which should be considered by individual beacons.

Assuming that an attack has been identified by a beacon within an open network then the source of the attack and the time of attack identification should be communicated through the network either to other beacons or to a head node/registry which propagates the alert around the network.

Given that an attack can come from multiple IPs running into an intractable number (more likely with DDos attacks), it will be necessary to store an IP list for an attack within a beacon and expose an endpoint returning IP addresses associated with a given attack for the head node or communicating beacon to set up a blacklist.

{
  "dos_attack": {
    "ip_address_endpoint": "",
    "no_of_requests": "",
    "start_time": "",
    "end_time": "",
    "threshold": ""
  },
  "reidentification_attack": {
    "query_set": [
      {
        "alternate_bases": "",
        "reference_name": "",
        "start": "",
        "exists": "",
        "ip_address": "",
        "timestamp": ""
      }
    ],
    "federated_id": "",
    "strategy": {
      "strategy_type": "",
      "strategy_description": "",
      "parameter_value": "",
      "parameter_description": ""
    }
  }
}