Toward Better Depth Lower Bounds: Strong Composition of XOR and a Random Function

Nikolai Chukhin,Alexander S. Kulikov,Ivan Mikhajlin

2024-10-14

Abstract:Proving formula depth lower bounds is a fundamental challenge in complexity theory, with the strongest known bound of $(3 - o(1))\log n$ established by Hastad over 25 years ago. The Karchmer--Raz--Wigderson (KRW) conjecture offers a promising approach to advance these bounds and separate P from NC$^{1}$. It suggests that the depth complexity of a function composition $f \diamond g$ approximates the sum of the depth complexities of $f$ and $g$. The Karchmer--Wigderson (KW) relation framework translates formula depth into communication complexity, restating the KRW conjecture as $\mathsf{CC}(\mathsf{KW}_f \diamond \mathsf{KW}_g) \approx \mathsf{CC}(\mathsf{KW}_f) + \mathsf{CC}(\mathsf{KW}_g)$. Prior work has confirmed the conjecture under various relaxations, often replacing one or both KW relations with the universal relation or constraining the communication game through strong composition. In this paper, we examine the strong composition $\mathsf{KW}_{\mathsf{XOR}} \circledast \mathsf{KW}_f$ of the parity function and a random Boolean function $f$. We prove that with probability $1-o(1)$, any protocol solving this composition requires at least $n^{3 - o(1)}$ leaves. This result establishes a depth lower bound of $(3 - o(1))\log n$, matching Hastad's bound, but is applicable to a broader class of inner functions, even when the outer function is simple. Though bounds for the strong composition do not translate directly to formula depth bounds, they usually help to analyze the standard composition (of the corresponding two functions) which is directly related to formula depth.

Computational Complexity

What problem does this paper attempt to address?

### The problem the paper attempts to solve

The paper aims to solve a fundamental challenge in complexity theory, that is, to prove formula depth lower bounds in Boolean formulas. The strongest known result so far was proposed by Håstad 25 years ago, that is, for any Boolean function \( f \), its minimum depth \( D(f) \) satisfies \( (3 - o(1)) \log n \). The Karchmer - Raz - Wigderson (KRW) conjecture provides a way to push these lower bounds forward and may separate \( P \) and \( NC^1 \). The conjecture states that the depth complexity of the composite function \( f \diamond g \) of two functions \( f \) and \( g \) is approximately equal to the sum of the depth complexities of \( f \) and \( g \).

### Main contributions

In this paper, the authors study the strong composition \( KWXOR_m \oplus KWf \), where \( XOR_m \) is the parity - check function and \( f \) is a random Boolean function. Specifically, they prove that when \( f \) is a balanced and hard - to - approximate function (which occurs with probability \( 1 - o(1) \)), any protocol for solving this composite problem requires at least \( n^{3 - o(1)} \) leaf nodes, where \( n = m \log m \). This result establishes a depth lower bound of \( (3 - o(1)) \log n \), which matches Håstad's result but applies to a broader class of inner functions, even when the outer function is simple.

### Proof methods

The authors use formal complexity measures to prove this lower bound. First, they apply Khrapchenko's method to show that after multiple communication steps, many instances of \( f \) remain unresolved. Subsequently, they switch to another formal complexity measure and prove that the remaining communication problem is at least as difficult as \( KWOR \oplus KWf \). This hybrid method not only achieves the required lower bound but also introduces a new technique for analyzing formula depth, which may provide guidance for future research in complexity theory.

### Key technical details

1. **Khrapchenko measure**: By tracking the progress of the protocol, ensure that even after multiple communication steps, a large number of instances of \( f \) still need to be resolved.
2. **Formal complexity measure**: Based on the Khrapchenko measure, switch to another formal complexity measure and prove that the remaining communication problem is at least as difficult as \( KWOR \oplus KWf \).
3. **Graph - theoretic method**: Construct a graph related to the protocol nodes and clean the graph by deleting light edges to ensure that the graph measure \( \psi \) does not decrease significantly.

### Conclusion

This paper proves a new depth lower bound by studying the communication complexity of the strong composition \( KWXOR_m \oplus KWf \), providing new tools and techniques for the development of complexity theory. This result not only verifies the validity of the KRW conjecture in some special cases but also provides new directions for future research.