DHCP spoofing remains one of the most common and disruptive Layer 2 attacks in local networks, where malicious packets impersonate legitimate DHCP servers to intercept or redirect traffic, much like social engineering scams that trick users into transferring funds.
In a new study, researchers evaluated three widely used switches under 90 simulated attacks to determine which hardware offers the strongest protection, with particular relevance for critical information infrastructure (CII) environments.
Devices Under Test
The team selected the Cisco 2960, a globally popular switch used for over a decade in banks, airports, and factories, featuring hardware-based security functions implemented in ASICs.
The Eltex MES1428 serves as a Russian alternative popular in government institutions and CII facilities, also relying on hardware chips for security features despite sanctions-related constraints on component sourcing.
The MikroTik CRS represents a budget option often found in regional providers, small offices, cafes, and occasionally larger enterprises due to cost savings, but with protection mechanisms handled by the central processor rather than dedicated hardware.
Experiment Setup
A pre-built university testbed allowed immediate testing. The legitimate DHCP server ran on ALT Linux, distributing addresses in the 192.168.1.100–200 range via gateway 192.168.1.1. The attacker used a Kali Linux laptop with a custom Scapy Python script to launch spoofing attempts, while the victim machine on ALT Linux requested configuration.
On Cisco and Eltex, DHCP snooping was enabled by designating trusted ports connected to the real server and treating all other ports as untrusted, causing the switch to drop unauthorized DHCP responses.
On MikroTik, protection was configured via Bridge Filter rules, a software-based approach executed by the CPU instead of specialized chips.
Results
Each device faced 30 attacks with protection active. Cisco blocked all 30 (100% effectiveness). Eltex blocked 28 (93.3%), and MikroTik blocked 24 (80%).
Detection times averaged 415 ms for Cisco, 650 ms for Eltex, and 1185 ms for MikroTik, with MikroTik showing high variance up to two seconds that allowed some attacks to succeed.
ROC analysis yielded AUC scores of 1.00 for Cisco, 0.98 for Eltex, and 0.90 for MikroTik. Latency increased by 18–22% across devices when protection was active, remaining acceptable for users.
CPU load rose modestly on Cisco (+3.1%) and Eltex (+5.1%), but MikroTik experienced nearly triple the increase (+8.9%, with peaks to +11%), potentially affecting other router functions such as VPN or internet speed.
Conclusion and Recommendations
The performance gap stems from hardware versus software implementation: Cisco and Eltex use dedicated ASICs, while MikroTik processes filters on the CPU.
For banks, large telecom operators, or CII objects, the study recommends Cisco or Eltex. For small offices, cafes, shops, or home networks, MikroTik may suffice but should be supplemented with Port Security, VLANs, and CPU monitoring rather than relying solely on Bridge Filter.
