Spaces:
Sleeping
Sleeping
File size: 11,061 Bytes
0ddfcb6 0df14c7 0ddfcb6 | 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65 66 67 68 69 70 71 72 73 74 75 76 77 78 79 80 81 82 83 84 85 86 87 88 89 90 91 92 93 94 95 96 97 98 99 100 101 102 103 104 105 106 107 108 109 110 111 112 113 114 115 116 117 118 119 120 121 122 123 124 125 126 127 128 129 130 131 132 133 134 135 136 137 138 139 140 141 142 143 144 145 146 147 148 149 150 151 152 153 154 155 156 157 158 159 160 161 162 163 164 165 166 167 168 169 170 171 172 173 174 175 176 177 178 179 180 181 182 183 184 185 186 187 188 189 190 191 192 193 194 195 196 197 198 199 200 201 202 203 204 205 206 207 208 209 210 211 212 213 214 215 216 217 218 219 220 221 222 223 224 225 226 227 228 229 230 231 232 233 234 235 236 237 238 239 240 241 242 243 244 245 246 247 248 249 250 251 252 253 254 255 256 257 258 259 260 261 262 263 264 265 266 267 268 269 270 271 272 273 274 275 276 277 278 279 280 281 282 283 284 285 286 287 288 289 290 291 292 293 294 295 296 297 298 299 300 301 302 303 304 305 306 307 308 309 310 311 312 313 314 315 316 317 318 319 320 321 322 323 324 325 326 327 328 329 330 331 332 333 334 335 336 337 338 339 340 341 | from fastapi import FastAPI
from pydantic import BaseModel
from typing import Optional
import uvicorn
import random
import math
from .models import Observation, Action, StepResponse
app = FastAPI(
title="CLAIRS Autonomous Defense Environment",
description="OpenEnv-compliant RL environment for IoT DDoS mitigation",
version="2.0.0",
)
class ResetRequest(BaseModel):
task_id: str = "task_1_easy"
class ActionPayload(BaseModel):
decision: str = "monitor"
ATTACK_PROFILES = {
"task_1_easy": {
"name": "Normal Traffic Monitoring",
"phases": [
{
"start": 0,
"end": 10,
"type": "normal",
"base_pps": 120,
"base_cpu": 10.0,
},
],
},
"task_2_medium": {
"name": "Volumetric DDoS Flood",
"phases": [
{
"start": 0,
"end": 2,
"type": "normal",
"base_pps": 200,
"base_cpu": 15.0,
},
{
"start": 2,
"end": 10,
"type": "attack_ramp",
"pps_start": 5000,
"pps_end": 50000,
"cpu_start": 55.0,
"cpu_end": 99.0,
},
],
},
"task_3_hard": {
"name": "Stealth Low-and-Slow DDoS",
"phases": [
{
"start": 0,
"end": 2,
"type": "normal",
"base_pps": 150,
"base_cpu": 12.0,
},
{
"start": 2,
"end": 10,
"type": "attack_ramp",
"pps_start": 2000,
"pps_end": 25000,
"cpu_start": 30.0,
"cpu_end": 75.0,
},
],
},
"task_4_expert": {
"name": "Multi-Wave APT Campaign",
"phases": [
{
"start": 0,
"end": 2,
"type": "normal",
"base_pps": 130,
"base_cpu": 11.0,
},
{
"start": 2,
"end": 5,
"type": "attack_ramp",
"pps_start": 4000,
"pps_end": 12000,
"cpu_start": 40.0,
"cpu_end": 60.0,
},
{
"start": 5,
"end": 7,
"type": "normal",
"base_pps": 180,
"base_cpu": 13.0,
},
{
"start": 7,
"end": 10,
"type": "attack_ramp",
"pps_start": 15000,
"pps_end": 45000,
"cpu_start": 70.0,
"cpu_end": 99.0,
},
],
},
}
class NetworkSimulator:
def __init__(self):
self.task_id = "task_1_easy"
self.step_count = 0
self.max_steps = 10
self.system_health = 100.0
self.current_pps = 100.0
self.current_cpu = 10.0
self.current_connections = 10
self.current_bandwidth = 1.0
self.current_memory = 30.0
self.false_positives = 0
self.attack_detected_step = None
self.cumulative_damage = 0.0
def reset(self, task_id: str) -> Observation:
self.task_id = task_id
self.step_count = 0
self.system_health = 100.0
self.false_positives = 0
self.attack_detected_step = None
self.cumulative_damage = 0.0
first_phase = ATTACK_PROFILES[task_id]["phases"][0]
noise = random.uniform(0.88, 1.12)
self.current_pps = first_phase["base_pps"] * noise
self.current_cpu = min(100.0, first_phase["base_cpu"] * random.uniform(0.9, 1.1))
self.current_connections = max(1, int(self.current_pps / 8 + random.randint(-5, 5)))
self.current_bandwidth = max(0.1, self.current_pps * 0.001 * random.uniform(0.8, 1.2))
self.current_memory = 25.0 + random.uniform(-3, 8)
return self._observation()
def step(self, action: str):
action = action.lower().strip()
if action not in ("monitor", "rate_limit", "block"):
action = "monitor"
reward = self._compute_reward(action)
self._advance_traffic(action)
self.step_count += 1
done = self.step_count >= self.max_steps
info = {
"mitigation_applied": action,
"is_attack_phase": self._is_attack(),
"attack_severity": round(self._severity(), 2),
"system_health": round(self.system_health, 1),
"false_positives": self.false_positives,
"cumulative_damage": round(self.cumulative_damage, 1),
}
return self._observation(), reward, done, info
def get_state(self) -> Observation:
return self._observation()
def _current_phase(self) -> dict:
for phase in ATTACK_PROFILES[self.task_id]["phases"]:
if phase["start"] <= self.step_count < phase["end"]:
return phase
return ATTACK_PROFILES[self.task_id]["phases"][-1]
def _is_attack(self) -> bool:
return self._current_phase()["type"] == "attack_ramp"
def _severity(self) -> float:
phase = self._current_phase()
if phase["type"] != "attack_ramp":
return 0.0
span = max(1, phase["end"] - phase["start"] - 1)
return min(1.0, (self.step_count - phase["start"]) / span)
def _advance_traffic(self, action: str):
phase = self._current_phase()
noise = random.uniform(0.88, 1.12)
mitigation = 1.0
if action == "block":
mitigation = 0.05 + random.uniform(0, 0.03)
elif action == "rate_limit":
mitigation = 0.35 + random.uniform(0, 0.08)
if phase["type"] == "normal":
target_pps = phase["base_pps"] * noise
target_cpu = phase["base_cpu"] * random.uniform(0.9, 1.1)
else:
span = max(1, phase["end"] - phase["start"] - 1)
progress = (self.step_count - phase["start"]) / span
ramp = min(1.0, progress ** 1.3)
raw_pps = phase["pps_start"] + (phase["pps_end"] - phase["pps_start"]) * ramp
raw_cpu = phase["cpu_start"] + (phase["cpu_end"] - phase["cpu_start"]) * ramp
target_pps = raw_pps * noise * mitigation
target_cpu = min(100.0, raw_cpu * noise * (0.3 + 0.7 * mitigation))
alpha = 0.7
self.current_pps = (1 - alpha) * self.current_pps + alpha * target_pps
self.current_cpu = (1 - alpha) * self.current_cpu + alpha * target_cpu
self.current_connections = max(1, int(self.current_pps / 8 + random.randint(-3, 3)))
self.current_bandwidth = max(0.1, self.current_pps * 0.001 * random.uniform(0.85, 1.15))
mem_delta = random.uniform(-2, 3)
if self._is_attack() and action == "monitor":
mem_delta += self._severity() * 4
self.current_memory = max(20.0, min(95.0, self.current_memory + mem_delta))
if self._is_attack() and action == "monitor":
dmg = self._severity() * random.uniform(3.0, 7.0)
self.system_health = max(0.0, self.system_health - dmg)
self.cumulative_damage += dmg
elif self._is_attack() and action == "rate_limit":
dmg = self._severity() * random.uniform(0.5, 2.0)
self.system_health = max(0.0, self.system_health - dmg)
self.cumulative_damage += dmg
else:
self.system_health = min(100.0, self.system_health + random.uniform(0.3, 1.0))
def _compute_reward(self, action: str) -> float:
is_attack = self._is_attack()
severity = self._severity()
reward = 0.50
if not is_attack:
if action == "monitor":
reward = 0.90 + random.uniform(0, 0.08)
elif action == "rate_limit":
reward = 0.25 + random.uniform(0, 0.08)
self.false_positives += 1
elif action == "block":
reward = 0.08 + random.uniform(0, 0.06)
self.false_positives += 1
else:
if severity > 0.6:
if action == "block":
reward = 0.88 + random.uniform(0, 0.09)
elif action == "rate_limit":
reward = 0.48 + random.uniform(0, 0.10)
else:
reward = 0.03 + random.uniform(0, 0.05)
elif severity > 0.2:
if action == "rate_limit":
reward = 0.85 + random.uniform(0, 0.09)
elif action == "block":
reward = 0.58 + random.uniform(0, 0.10)
else:
reward = 0.05 + random.uniform(0, 0.07)
else:
if action in ("rate_limit", "block"):
reward = 0.78 + random.uniform(0, 0.10)
else:
reward = 0.10 + random.uniform(0, 0.08)
if self.attack_detected_step is None and action in ("rate_limit", "block"):
self.attack_detected_step = self.step_count
if self.step_count <= 3:
reward = min(0.99, reward + 0.04)
if self.task_id == "task_3_hard" and is_attack:
if action == "rate_limit":
reward = min(0.99, reward + 0.04)
elif action == "block" and severity < 0.5:
reward = max(0.01, reward - 0.08)
if self.system_health > 70:
reward = min(0.99, reward + 0.02)
return round(max(0.01, min(0.99, reward)), 4)
def _observation(self) -> Observation:
return Observation(
cpu_usage_percent=round(self.current_cpu, 2),
packet_rate_pps=round(self.current_pps, 2),
active_connections=max(0, self.current_connections),
bandwidth_mbps=round(self.current_bandwidth, 2),
memory_usage_percent=round(self.current_memory, 2),
system_health=round(self.system_health, 2),
)
simulator = NetworkSimulator()
@app.post("/reset")
def reset(req: Optional[ResetRequest] = None):
task_id = req.task_id if req else "task_1_easy"
if task_id not in ATTACK_PROFILES:
task_id = "task_1_easy"
obs = simulator.reset(task_id)
return obs.model_dump()
@app.post("/step", response_model=StepResponse)
def step(payload: Optional[ActionPayload] = None):
action = payload.decision.lower() if payload else "monitor"
obs, reward, done, info = simulator.step(action)
return StepResponse(observation=obs, reward=reward, done=done, info=info)
@app.get("/state", response_model=Observation)
def state():
return simulator.get_state()
@app.get("/health")
def health():
return {"status": "ok"}
def main():
uvicorn.run("server.app:app", host="0.0.0.0", port=7860)
if __name__ == "__main__":
main()
|