移动设备间的高效安全协议：舞动信号（TDS）

60 浏览量更新于2024-08-26 收藏 684KB PDF 举报

本文主要探讨了移动设备之间即时且强大身份验证与密钥协议的设计与实现，针对新兴的移动应用如物联网和移动社交网络，设备间的安全通信是至关重要的。现有的解决方案往往依赖于物理近距离接触，利用无线电环境作为身份认证的证据，并共享共同的秘密来协商生成同一密钥。然而，这些方法存在显著的局限性，如秘钥生成速度慢（可能需要几分钟才能建立一个256位的密钥）以及较高的错误率。作者团队，来自西安交通大学、加利福尼亚大学圣克鲁斯分校、南京大学和中国科学技术大学的研究人员，提出了名为"The Dancing Signals (TDS)"的全新协议。TDS旨在解决现有方案中的问题，通过创新的技术手段，实现在无需物理接近的情况下，提高密钥生成的效率和安全性。TDS协议可能采用了先进的加密算法、无线信号特征分析或利用设备间的同步行为来增强信任度，从而加快身份验证和密钥协商过程。 TDS可能包括以下几个关键特性： 1. **远程身份验证**：利用无线信号的特性或者设备特有的行为模式，即使在远处也能进行准确的身份验证，减少对物理位置的依赖。 2. **高效密钥交换**：通过优化的算法设计，可能使用零知识证明、多因素认证或动态密钥生成技术，显著提升密钥生成速度，使得建立高强度密钥所需时间显著缩短。 3. **错误容错机制**：考虑到无线环境的不稳定性，TDS可能会包含错误检测和纠正机制，以确保在复杂环境中仍能维持高可靠性。 4. **适应移动环境**：TDS可能考虑到了移动设备的移动性和连接变化，能够实时调整并保持安全连接，适应不断变化的无线环境。 5. **安全性保障**：通过采用现代密码学理论和安全模型，确保协议在对抗各种攻击（如重放攻击、中间人攻击等）时具有足够的防护能力。总结来说，The Dancing Signals (TDS)是一项旨在推动移动设备间即时、高效且安全通信的革新性研究，它为物联网和移动社交网络这类应用提供了重要的安全保障，有望在未来改变移动设备身份验证和密钥管理的现状。

−0.2

0.2

0.4

0.6

0.8

Distance

Correlation coefficien

Indoor

Outdoor









Figure 1: Pearson corre-

lation coeﬃcient decreas-

es with growing distance

2



3

4

Mismatch rate

0.

0.

0.

0.

'LVFDUGHG rate

0.2

0.4

0.6

0.8

MismatchUDWH

'LVFDUGHG UDWH

$EDQGRQHG=RQH

Figure 2: Dilemma of mis-

match rate and discarded

rate

0 5 10 15 20 25 30 35 40

# of Error %its

100

150

200

250

# of Check 7imes

RISDVVHV 5

RISDVVHV 

RISDVVHV 

Figure 3: Mismatched

bits cause high cost in er-

ror correction

0 5 10 15 20 25 30 35 40

# of Error Bits

0.2

0.4

0.6

0.8

Recall Rate

5 times

6 times

7 times

Figure 4: Information

reconciliation fails more

seriously with increasing

error bits

Eve know about the key. We have the following security

requirements.

• Authenticity. A device needs to ensure that it is mak-

ing key agreement with other legitimate devices. Any

spoofer will be detected.

• Conﬁdentiality. Any information of the key should not

be exposed to Eve.

• Integrity. The key should be consistent at all legiti-

mate devices.

We do not consider availability in this work. If Eve block

all WiFi signals, devices may not agree on any key.

2.2 Feasibility of CSI-based key generation

In this paper, we use CSI as the proof of authenticity and

source of common secret information. The intuition of us-

ing CSI is that it is a unique and correlated measurement

for devices around a particular physical location. CSI mea-

surements at diﬀerent devices are rapidly de-correlated with

distance between them. In addition, CSI is unpredictable

due to its random property caused by the multipath eﬀect

of signal propagation. CSI is a much richer source of secret

information than the one of RSS, because it contains the

information of 56 subcarriers in each measurement sample.

We demonstrate the properties using experimental vali-

dation. We use two laptop computers Alice and Bob, each

equipped with COTS wireless NIC model Intel 5300 oper-

ating in the 802.11n 2.4GHz channel.

They then collect

measurement results of the CSI amplitude values indepen-

dently from a public WiFi while varying their distance from

0.1λ to 2.5λ, where the wavelength λ =12.5cm for 2.4GHz.

Figure 1 shows the Pearson product-moment correlation co-

eﬃcient of the CSI samples from the two devices. We found

that when the distance is smaller than 0.5λ, the samples are

highly correlated. The correlation drops quickly with the

distance growth. When the distance > 2λ, the samples are

uncorrelated.

The above properties of CSI are important for device au-

thentication and key agreement. Only if the samples from

diﬀerent devices nearby are similar, CSI can be a proof of

We use laptops for the ease of programming. The method

can be applied to any devices with 802.11 NICs.

physical proximity. Only if the samples are rapidly de-

correlated with distance between the devices, CSI can be

a common secure information.

2.3 Challenges of CSI-based key generation

Suppose two devices, Alice and Bob both listen to a public

WiFi source. For each of them, the CSI amplitude value h(t)

at time t can be directly obtained from an existing API of

the Intel 5300 network card. To extract secret information

from two similar measurements of CSI amplitude values, a

simple approach is to determine a cut-oﬀ amplitude level

h anduse0torepresentsamplessmallerthanh and 1 to

represent samples larger than h. For example, h can be 0.5

for CSI amplitude varying in [0, 1]. This method is called

reciprocal quantization.

Reciprocal quantization may cause mismatched bits at t-

wo devices. For example, if Alice gets a CSI value 0.53 for

a particular bit, Bob gets 0.48, and the cut-oﬀ is 0.5, then

they will have a diﬀerent bit. To reduce these mismatched

bits, existing quantization methods often use an abandoned

zone. For example, if the abandoned zone is [0.4, 0.6], then

only if a CSI value is less than 0.4 (or larger than 0.6), it

canbeconvertedtoa0-bit(or1-bit).

Selecting the size of the abandoned zone is a dilemma: if

the zone is small, mismatched bits still occur; if the zone

is large, too many CSI samples will be discarded, slowing

secret bit generation. Figure 2 shows the bit mismatch rate

versus the discarded bit rate by varying the abandoned zone

from 0 to 4σ, where σ is the standard deviation of the Gaus-

sian noise. We ﬁnd that when the abandoned zone is smaller

than σ, the discarded rate is low but it causes more than 10%

mismatched bits. When the zone is large, e.g.,3σ,themis-

match rate is negligible but more than 80% samples will be

discarded. To further demonstrate the harm of mismatched

bits, we use an existing method, information reconciliation

[3] [8], to ﬁx mismatched bits by iterative parity checks.

Figure 3 shows that the rounds of parity checks increase sig-

niﬁcantly with growing mismatched bits, for generation of a

256-bit key. For 20 mismatched bits, it requires more than

70-150 parity check bits to correct them. Besides tremen-

dous communication and time cost, the number of secret

bits is also reduced from 256 to < 150 due to privacy am-

pliﬁcation [17]. Additionally, information reconciliation is a

618

剩余11页未读，继续阅读

weixin_38672840

粉丝: 9
资源: 893

移动设备间的高效安全协议：舞动信号（TDS）

面向云服务的安全高效无证书聚合签名车联网认证密钥协商协议.pdf

电信设备-基于即时通讯的安全语音通讯方法及装置.zip

使用椭圆曲线密码学的会话发起协议的匿名两因素认证密钥协商方案

电信设备-接收字符短信息的移动通信装置.zip

wap协议

电子政务-一种短信密钥控制的防盗门电控系统.zip

qq协议文档

sms_移动短信_

Java支付宝即时到账接口demo源码下载

基于python的安全即时通讯系统.zip

最新资源