One-line positioning
A protocol fuzz testing platform using graphical state-machine orchestration and 22 mutation strategies to construct abnormal protocol packets, with AI-generated test cases and large-scale test case fuzzing to validate device protocol stack robustness.
Product overview
XproFuzzing addresses protocol fuzz testing for network equipment through a layered, modular architecture covering the entire “user operation—core execution—performance support” process. Its five layers (physical, DPDK, protocol encoding/decoding module, application and configuration management) decouple hardware and software. Testers can build element topologies and protocol state flows through graphical drag-and-drop operations, orchestrating complex test scenarios without specialist programming skills.
Test dimensions
- Protocol conformance testing: validate a protocol against standard specifications or custom rules
- Protocol security testing: use abnormal data and interaction-logic attacks to discover buffer overflow, denial-of-service, route hijacking and other security risks
Core capabilities
- Reverse generation of graphical state machines from PCAP: import PCAP captures to parse protocol interaction logic and generate visual state machines, reducing the difficulty of testing unknown protocols. For example, PPPoE captures generate Discovery and Session interaction flows with PADI / PADO / PADR / PADS state nodes
- Graphically programmable protocol state flows: drag state nodes and set transition conditions to build interaction flows; script message parameters for each node, such as source port, destination port, sequence number and control bits; built-in syntax validation reports configuration errors in real time
- Protocol state flow template library:
- Basic protocol templates: TCP connection establishment/teardown, HTTP request/response and ICMP echo request/reply
- Scenario templates: BGP route updates, PPPoE dial-up connections and NGAP initial UE message exchanges
- Attack templates: TCP SYN Flood, BGP route hijacking and PPP authentication bypass
- Automatic test case generation: template-based generation (variable parameter replacement) and mutation-based generation (multi-dimensional mutation for abnormal and compliant cases), covering normal interactions, abnormal scenarios and boundary conditions, with traceability and automatic updates as protocol rules change
- Custom protocol state flows: create field definitions, state nodes and interaction logic for private / new protocols from scratch, with reusable templates and import/export
Protocol coverage (46 protocols)
| Layer | Protocols |
|---|---|
| Link layer and Layer 2 control (15) | PPPoE, LLDP, LLDP (IPv6), STP, RSTP, ESTP, MSTP, L2TPv2, L2TPv3, PTP, gPTP, LACP, OpenFlow, CFM, EFM 802.3ah |
| Network layer (2) | IPv4, IPv6 |
| Network and control layer (15) | ARP, ICMP, ICMPv6, IGMP (V1/V2), IGMP (V3), GRE, RIP, VRRP, VRRP (IPv6), BFD, IS-IS, BGP, BGP4+, MPLS, RSVP |
| Transport layer (3) | TCP, UDP, QUIC |
| Application layer (8) | DHCP, DNS, FTP, HTTP, NTP, NTP (IPv6), DoIP, SOME/IP |
| Security layer (3) | IPsec-AH, IPsec-ESP, TLS |
Mutation strategies
| Category | Strategies |
|---|---|
| Data value mutation | random, bound, special, sequence, variance, length, length_variance, case |
| Data encoding mutation | value_flip, value_slider, value_change, value_expand, value_expand_variance |
| Field encoding mutation | flip, slider, change, expand, expand_variance, repeat |
| Field byte-stream compilation | present |
| Data byte-stream encoding | reorder, swap |
Test modes: “fuzz testing” and “exhaustive fuzz testing”. Fuzz testing requires mutation settings on message template fields. Exhaustive fuzz testing runs through all currently associated supported mutation types after the iteration count is set.
Protocol message definition DSL (domain-specific language)
- Standard and custom protocol models, with built-in examples for ARP, TCP, UDP, ICMP, IPv4, IPv6, DHCP, BGP and PPPoE
- Data types: u8/u16/u32/u64, i8/i16/i32/i64, str, dot_str, bytes, bytesn, bitset, bcd, ipv4_address, ipv6_address and more, plus compound types such as struct, vec, map, choice and nullable
- Field attribute modifiers: COUNT, COUNT_RANGE, VEC, CHOICE, LV, TLV, TV/TVS/TLVE/TLVEE/AVP, PADDING, LINK, DERIVE, SPAN_LEN, BOUND, RECV_COND and more, describing fixed/variable-length fields, conditional branches, nesting, length derivation and parsing boundaries
- Built-in functions: l4cks_encode (L4 checksum completion), ip4cks_encode (IPv4 header checksum), dynlen_encode / dynint_encode (dynamic length completion), msg_data_len (received message length), rand_int / rand_str / rand_bytes (random value generation)
- Input methods: DSL import (cfg files), PCAP import and associated JSON import
State monitoring
- Host reachability: configure the target device IP to check connectivity
- TCP port connectivity: configure the target device IP and TCP port to check connectivity
Performance support architecture
- NET component: packet transmission, reception and forwarding in conventional networks, for low-to-medium traffic and normal concurrency
- DPDK component: a DPDK-based user-space protocol stack avoiding kernel-space transport delay, for high traffic and high concurrency
- Hot loading / unloading of protocol plugins and test components to extend supported protocols without restarting
Deployment requirements
| Item | Requirement |
|---|---|
| Operating system | Ubuntu 24.04, kernel 6.8.0-41-generic |
| Resources | At least 5GB memory and 4 CPU cores |
| NICs | VMXNET3 and e1000e virtual NICs; Mellanox NICs bound to the kernel |
| NIC binding | DPDK driver binding supported |
| Management | B/S architecture, Chrome and other mainstream browsers, with HTTPS |
| Licensing | Hardware UKEY / software lock, with local license updates and migration |
