We investigate the benefits of using multiple channels in wireless networks, under the full-duplex multi-packet reception model of communication. The main question we address is the following: Is a speedup linear in the number of channels available, for some interesting communication primitive? We provide a positive answer to this interrogative for the Information Exchange Problem, in which up to k arbitrary nodes have information they intend to share with the entire network. In particular, we give non-adaptive deterministic protocols for both the scenario in which the channels provide the transmitting stations with the feedback on whether their transmissions have been successful and for the scenario in which channels provide no such feedback. To this aim, we devise and exploit a new combinatorial structure that generalizes well known combinatorial tools, widely used in the area of data-exchange in multiple-access channels (i.e., strongly selective families, selectors, and related mathematical objects). For our new combinatorial structures we provide both existential results and randomized algorithms to generate them. We also prove non-existence results showing that our protocol for the model with feedback is optimal, whereas that for the no-feedback scenario uses a number of time slots that exceeds the lower bound by a $\log k$ factor. Leveraging on properties of error correcting codes, we show that for an infinite set of the relevant parameters n and k, one can construct our combinatorial structure for the no-feedback scenario in polynomial time and of minimum length.

我们研究了在全双工通信的多分组接收模型下在无线网络中使用多个信道的好处。我们要解决的主要问题如下：对于一些有趣的通信原语，提速是否在可用通道数上呈线性关系？对于这一信息查询问题，我们提供了肯定的答案，其中最多k任意节点具有它们打算与整个网络共享的信息。特别地，对于信道向发送站提供关于其传输是否成功的反馈的情况以及信道不提供这种反馈的情况，我们都给出非自适应确定性协议。为此，我们设计并开发了一种新的组合结构，该结构将众所周知的组合工具进行了概括，该组合工具广泛用于多访问通道（即强选择族，选择器和相关数学对象）的数据交换领域。对于我们的新组合结构，我们提供了存在性结果和生成它们的随机算法。我们还证明了不存在的结果，表明我们的带反馈模型的协议是最优的，$\mathrm{日志}ķ$因子。利用纠错码的特性，我们表明，对于相关参数n和k的无穷集，可以在多项式时间内以最小长度构造无反馈情况的组合结构。