Ag Versus Wolves
Congratulations to Wolves Team.The ninth champion!
Reading The Paper
It is a pity not to post the emoji.I want to say who in the end understands?
If you are interesting in it,please look through it presumably.

5.3 Guided Sequence Generation
Navigated by the SSTG, we devise heuristics to efficientlyexplore unknownstate space and facilitate comprehensivetraversal. This module outputs the packet pattern sequenceand passes it to the packet instantiationsub-module as a guideline for packet sequence generation.
Algorithm 1: Guided packet sequence generation
Input: SSTG: System State Tracking Graph
Input: PT : transition priority
Output: ϕ: packet pattern sequence 

1
2
3
4
5
6
7
8
9
10
11
12
S ← TRAVERSE(SSTG)  
f S is not empty then  
q ← SELECTONE(S)  
ϕ ← ENTERANDAPPEND(SSTG,q)  
else  
ϕ ← ∅  
q ← GETINITIALSTATE(SSTG)  
while HASSUCCESSOR(q) do  
τ ← SELECTTRANSITION(SSTG,PT ,q)  
ϕ ← ϕ∪ GETPACKETPATTERN(τ)  
q ← TRANSITION(SSTG,q, τ)  
return ϕ

Algorithm 1 provides an overview of the process: first, weattempt to stress test each SSTG state with diverse inputs by utilizing the proposed mutation operators (Lines 1-4). Specifically, for a SUT State h C(α)|S(β)i ,although the state output α is “deterministic” (the corresponding concrete packets of each execution may vary in some details) such that BLEEM receives the same packet type α as the mutation basis, we can select a different mutation operator σ to construct a different packet pattern α ⊕ σ as test inputs. To this end, we first check whether there is a state q that can be further exercised by other kinds of packet patterns that have not beenapplied (Line 1). If so, we construct a packet pattern sequence ϕ that can reach q and then append the desired packet pattern after ϕ (Line 4). For example, for the SSTG in Figure 6 with Σ 0 = {σ◦,σP,σS}, if stressing q1 with a new packet pattern that combines q1’s output b with an unused mutation operator σP, i.e., b⊕σP, then we can construct such a packet pattern sequence: [a ⊕ σ◦,b ⊕ σP]. Second, after all the SUT states have been exercised by all kinds of packet patterns that are available, we attempt to facilitate comprehensive traversal of the implemented SSTG by steering towards low-density regions. We start from the initial state (Line 7) and then run in a loop until reaching an end state (Line 8): in each step, we select the one with the highest priority among the available transitions of the corresponding state (Line 9), record the corresponding packet pattern labeled on it (Line 10), and take this transition (Line 11). It is worth noting that even if we have once applied a packet pattern P to stress a state S and failed to discover new behaviors, it is also worthwhile trying P on S later because, as mentioned above, the set of the packet(s) that P can produce is typically infinite.