/
DDPG.py
179 lines (123 loc) · 6.46 KB
/
DDPG.py
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
import copy
import torch
import torch.nn as nn
import torch.nn.functional as F
import numpy as np
# Implementation of the Deep Deterministic Policy Gradient algorithm (DDPG)
# Paper: https://arxiv.org/abs/1509.02971
class Actor(nn.Module):
def __init__(self, state_dim, action_dim, M, N, K, power_t, device, max_action=1):
super(Actor, self).__init__()
hidden_dim = 1 if state_dim == 0 else 2 ** (state_dim - 1).bit_length()
self.device = device
self.M = M
self.N = N
self.K = K
self.power_t = power_t
self.l1 = nn.Linear(state_dim, hidden_dim)
self.l2 = nn.Linear(hidden_dim, hidden_dim)
self.l3 = nn.Linear(hidden_dim, action_dim)
self.bn1 = nn.BatchNorm1d(hidden_dim)
self.bn2 = nn.BatchNorm1d(hidden_dim)
self.max_action = max_action
def compute_power(self, a):
# Normalize the power
G_real = a[:, :self.M ** 2].cpu().data.numpy()
G_imag = a[:, self.M ** 2:2 * self.M ** 2].cpu().data.numpy()
G = G_real.reshape(G_real.shape[0], self.M, self.K) + 1j * G_imag.reshape(G_imag.shape[0], self.M, self.K)
GG_H = np.matmul(G, np.transpose(G.conj(), (0, 2, 1)))
current_power_t = torch.sqrt(torch.from_numpy(np.real(np.trace(GG_H, axis1=1, axis2=2)))).reshape(-1, 1).to(self.device)
return current_power_t
def compute_phase(self, a):
# Normalize the phase matrix
Phi_real = a[:, -2 * self.N:-self.N].detach()
Phi_imag = a[:, -self.N:].detach()
return torch.sum(torch.abs(Phi_real), dim=1).reshape(-1, 1) * np.sqrt(2), torch.sum(torch.abs(Phi_imag), dim=1).reshape(-1, 1) * np.sqrt(2)
def forward(self, state):
a = torch.tanh(self.l1(state.float()))
# Apply batch normalization to the each hidden layer's input
a = self.bn1(a)
a = torch.tanh(self.l2(a))
a = self.bn2(a)
a = torch.tanh(self.l3(a))
# Normalize the transmission power and phase matrix
current_power_t = self.compute_power(a.detach()).expand(-1, 2 * self.M ** 2) / np.sqrt(self.power_t)
real_normal, imag_normal = self.compute_phase(a.detach())
real_normal = real_normal.expand(-1, self.N)
imag_normal = imag_normal.expand(-1, self.N)
division_term = torch.cat([current_power_t, real_normal, imag_normal], dim=1)
return self.max_action * a / division_term
class Critic(nn.Module):
def __init__(self, state_dim, action_dim):
super(Critic, self).__init__()
hidden_dim = 1 if (state_dim + action_dim) == 0 else 2 ** ((state_dim + action_dim) - 1).bit_length()
self.l1 = nn.Linear(state_dim, hidden_dim)
self.l2 = nn.Linear(hidden_dim + action_dim, hidden_dim)
self.l3 = nn.Linear(hidden_dim, 1)
self.bn1 = nn.BatchNorm1d(hidden_dim)
def forward(self, state, action):
q = torch.tanh(self.l1(state.float()))
q = self.bn1(q)
q = torch.tanh(self.l2(torch.cat([q, action], 1)))
q = self.l3(q)
return q
class DDPG(object):
def __init__(self, state_dim, action_dim, M, N, K, power_t, max_action, actor_lr, critic_lr, actor_decay, critic_decay, device, discount=0.99, tau=0.001):
self.device = device
powert_t_W = 10 ** (power_t / 10)
# Initialize actor networks and optimizer
self.actor = Actor(state_dim, action_dim, M, N, K, powert_t_W, max_action=max_action, device=device).to(self.device)
self.actor_target = copy.deepcopy(self.actor)
self.actor_optimizer = torch.optim.Adam(self.actor.parameters(), lr=actor_lr, weight_decay=actor_decay)
# Initialize critic networks and optimizer
self.critic = Critic(state_dim, action_dim).to(self.device)
self.critic_target = copy.deepcopy(self.critic)
self.critic_optimizer = torch.optim.Adam(self.critic.parameters(), lr=critic_lr, weight_decay=critic_decay)
# Initialize the discount and target update rated
self.discount = discount
self.tau = tau
def select_action(self, state):
self.actor.eval()
state = torch.FloatTensor(state.reshape(1, -1)).to(self.device)
action = self.actor(state).cpu().data.numpy().flatten().reshape(1, -1)
return action
def update_parameters(self, replay_buffer, batch_size=16):
self.actor.train()
# Sample from the experience replay buffer
state, action, next_state, reward, not_done = replay_buffer.sample(batch_size)
# Compute the target Q-value
target_Q = self.critic_target(next_state, self.actor_target(next_state))
target_Q = reward + (not_done * self.discount * target_Q).detach()
# Get the current Q-value estimate
current_Q = self.critic(state, action)
# Compute the critic loss
critic_loss = F.mse_loss(current_Q, target_Q)
# Optimize the critic
self.critic_optimizer.zero_grad()
critic_loss.backward()
self.critic_optimizer.step()
# Compute the actor loss
actor_loss = -self.critic(state, self.actor(state)).mean()
# Optimize the actor
self.actor_optimizer.zero_grad()
actor_loss.backward()
self.actor_optimizer.step()
# Soft update the target networks
for param, target_param in zip(self.critic.parameters(), self.critic_target.parameters()):
target_param.data.copy_(self.tau * param.data + (1 - self.tau) * target_param.data)
for param, target_param in zip(self.actor.parameters(), self.actor_target.parameters()):
target_param.data.copy_(self.tau * param.data + (1 - self.tau) * target_param.data)
# Save the model parameters
def save(self, file_name):
torch.save(self.critic.state_dict(), file_name + "_critic")
torch.save(self.critic_optimizer.state_dict(), file_name + "_critic_optimizer")
torch.save(self.actor.state_dict(), file_name + "_actor")
torch.save(self.actor_optimizer.state_dict(), file_name + "_actor_optimizer")
# Load the model parameters
def load(self, file_name):
self.critic.load_state_dict(torch.load(file_name + "_critic"))
self.critic_optimizer.load_state_dict(torch.load(file_name + "_critic_optimizer"))
self.critic_target = copy.deepcopy(self.critic)
self.actor.load_state_dict(torch.load(file_name + "_actor"))
self.actor_optimizer.load_state_dict(torch.load(file_name + "_actor_optimizer"))
self.actor_target = copy.deepcopy(self.actor)